An Introduction To Bilingualism

ftnHnrr tVift Innnrii^^fTn P'

di An Introduction to
at
ur BILINGUALISM
lich
Charlotte Hoffmann

i uui iiwHim" > i. C l

p^pi|in||Hpariz WK'
•m vi.bru^r vor s cgnc

bee n kjgpi
s c, o u i i u y fo i pIong||iic U
id fcel^Rrai.igcr any^| nd I d<ftnio|
dng Hiânûaye]
Lino sierr™| un Ii||p3 i^
disD ■feAs cn îxiranjcro,“
11
P»>usiq: ad ms, sc,g^jioia .
"Als sicliûi hierfioh iiaba- rrdch

an 3pra|j|e, Te^jj^wieni ijfti Lej^n»(eis«''
ijiic uno pcx

disiinto de l^^dernas en un pais cxlra^*.ro.
imillllklc^lby a^jjUlBilîîas, sc
accnio
n d ci^
Ksprae frro l.,.l]x/ siA'cj
1 mil dcr MuiJ
\I/fV
Longman Library
LONGMAN LINGUISTICS LIBRARY
AN INTRODUCTION TO BILINGUALISM
LONGMAN LINGUISTICS LIBRARY
General editors Pidgin and Creole Languages

R. H. Robins, University of London SUZANNE ROMAINE
Martin Harris, University of Essex
General Linguistics
An Introductory Survey
Fourth Edition
R. H. ROBINS
Structural Aspects of Language A History of English Phonology

Change CHARLES JONES
JAMES M. ANDERSON
Generative and Non-linear
Text and Context Phonology
Explorations in the Semantics and JAQUES DURAND
Pragmatics of the Discourse
TEUN A. VAN DIJK Modality and the English
Models
Introduction to Text Linguistics Second Edition
ROBERT-ALAIN DE BEAUGRANDE F. R. PALMER
AND WOLFGANG ULRICH
DRESSIER Semiotics and Linguistics
YISHAI TOBIN
Spoken Discourse
A Model, for Analysis Multilingualism in the British
WILLIS EDMONDSON Isles I: the Older Mother
Tongues and Europe
Psycholinguistics EDITED BY SAFDER ALLADINA
Language, Mind, and World AND VIV EDWARDS
DANNY D. STEINBERG
Multilingualism in the British
Dialectology Isles II: Africa, Asia and the Middle
W. N. FRANCIS East
EDITED BY SAFDER ALLADINA
Principles of Pragmatics AND VIV EDWARDS
GEOFFREY N. LEECH
Dialects of English
Generative Grammar Studies in Grammatical Variation
GEOFFREY HORROCKS EDITED BY PETER TRUDGILL
AND J. K. CHAMBERS
Norms of Language
Theoretical and Practical Aspects A Short History of Linguistics
RENATE BARTSCH Third Edition
R. H. ROBINS
The English Verb
Second Edition An Introduction to Bilingualism
F. R. PALMER CHARLOTTE HOFFMANN
An Introduction to
Bilingualism
Charlotte Hoffmann
HUH
mmm
Lxjngman
London and New York
Longman Group UK Limited,
Longman House, Burnt Mill, Harlow,
Essex CM20 2JE England
and Associated Companies throughout the world.
Published in the United States of America

by Longman Inc., New York
© Longman Group UK Limited 1991
All rights reserved; no part of this publication may be

reproduced, stored in a retrieval system, or transmitted in
any form or by any means, electronic, mechanical,
photocopying, recording, or otherwise without either
the prior written permission of the Publishers or a licence
permitting restricted copying in the United Kingdom issued
by the Copyright Licensing Agency Ltd., 33-34 Alfred Place,
London, WC1E 7DP.
First published 1991
British Library Cataioguing in Pubiication Data

Hoffmann, Charlotte
An introduction to bilingualism. (Longman linguistics library).
1. Bilingualism
I. Title
404.2
ISBN 0-582-07920-9
ISBN 0-582-29143-7 pbk
Library of Congress Cataioging in Pubiication Data

Hoffmann, Charlotte, 1947-
An introduction to bilingualism/Charlotte Hoffmann.
p. cm. - (Longman linguistics library)
Includes bibliographical references and index.
ISBN 0-582-07920-9 (cased). - ISBN 0-582-29143-7 (paper)
1. Bilingualism. I. Title.
II. Series
P115.H54 1991
404’.2-dc20
90-6584
CIP
Set in Linotron 200 10/12pt Times
Produced by Longman Singapore Publishers (Pte) Ltd

Printed in Singapore
Contents
List of maps x
Preface xi
Introduction X
PART I PSYCHOLINGUISTIC ASPECTS OF BILINGUALISM
1 Individual bilingualism 13
1.1 Societal and individual bilingualism 13
1.2 Describing bilingualism 14
1.2.1 Some definitions of bilingualism 15
1.2.2 Some types of bilinguals 16
1.2.3 Factors taken into account when describing
bilingualism 1g
1.3 Semilingualism 27
1.4 Biculturalism 28
1.5 A bilingual profile 31
2 The study of bilingual children 33

2.1 Early and late bilingualism 33
2.1.1 Is early bilingualism ‘better’ than late
bilingualism? 35
2.1.2 The Language Acquisition Device (LAD) and
the Critical Period Hypothesis 36
2.1.3 Children are not necessarily better
bilinguals 37
2.2 How does the child become bilingual? 38
2.2.1 Some examples 38
2.2.2 Bilingual patterns 40
2.3 Types of bilinguals 46
2.4 Case studies of bilingual children 48
VI CONTENTS
3 Patterns of bilingual language acquisition 55

3.1 The description of bilingual language acquisition 55
3.2 Phonological development 56
3.3 Semantic and lexical development 60
3.3.1 General acquisition processes 60
3.3.2 Lexical development in bilingual children 62
3.3.3 Acquisition of equivalents 63
3.3.4 The semantic load 65
3.4 Development of grammar 66
3.4.1 Morphological development 66
3.4.2 Syntactic development 69
4 Aspects of bilingual competence 74

4.1 The bilingual’s linguistic competence 74
4.2 Early bilingualism: one language or two? 75
4.2.1 The unitary language system hypothesis 75
4.2.2 The separate or independent development
hypothesis 76
4.2.3 An assesment of the two hypotheses 77
4.3 Language awareness and differentiation 79
4.3.1 Manifestations of language awareness 80
4.3.2 How does the child keep the languages
apart? 85
4.3.3 The importance of context 86
4.4 Language choice in children 88
4.4.1 Language choice in monolinguals and
bilinguals 88
4.4.2 Determinants of language choice in
bilingual children 89
5 Features of bilingual speech 94

5.1 Interference 95
5.1.1 Interference at the phonological level 96
5.1.2 Interference at the grammatical level 97
5.1.3 Interference at the lexical level 99
5.1.4 Interference in spelling 100
5.2 Borrowing 101
5.3 Individual creations 103
5.4 Mixing 103
CONTENTS vii
5.5 Code-switching 109

5.5.1 What is code-switching? 110
5.5.2 What types of code-switches are there? 111
5.5.3 Who code-switches? 113
5.5.4 Can code-switching occur anywhere? 114
5.5.5 Why do bilinguals code-switch? 115
6 Cognitive and educational aspects of bilingualism 118

6.1 Contradictory research findings 118
6.1.1 Bilingualism and cognitive functioning 120
6.1.2 Overview of research on bilingualism and cognitive

functioning 121
6.2 Linguistic competence, cognitive functioning and the
education of minority children 127
6.2.1 Semilingualism - ‘BICS’ and ‘CALP’ 127
6.2.2 The Balance and the Think Tank models 129
6.2.3 The Threshold Theory 130
6.3 Summary of the issues 131
7 Sociocultural aspects of bilingualism 136

7.1 Issues raised in the discussion of bilingualism 136
7.2 Views of bilingualism 137
7.3 Psychological aspects 141
7.3.1 Stuttering 141
7.3.2 Personality development 142
7.4 Linguistic aspects 148
7.4.1 Assessing bilinguals 148
7.4.2 Linguistic competence as the object of
assessment 149
7.4.3 Measuring bilingual language skills 151
7.4.4 The problems of assessing bilingual
proficiency 152
PART II SOCIOLINGUISTIC ASPECTS OF BILINGUALISM
8 Societal multilingualism 157

8.1 Multilingualism in society 157
8.2 Factors contributing towards societal
multilingualism 158
viii CONTENTS
8.2.1 Historical factors 158

8.2.2 Contemporary factors 160
8.3 Patterns of societal multilingualism 163
8.3.1 Determinants of multilingual patterns 163
8.3.2 Basic types of societal multilingualism 164
8.3.3 Diglossia 166
8.3.4 Examples of multilingual patterns in
contemporary Europe 169
8.4 The unstable nature of multilingual situations 173
9 Language choice, language maintenance and

language shift 175
9.1 Language choice 175
9.1.1 Domain analysis: person, place and topic 177
9.1.2 The socio-psychological approach: linguistic
accommodation 180
9.1.3 Anthropological methods of observing language
choice 182
9.2 Language maintenance 185
9.3 Language shift and its causes 186
9.3.1 Migration 189
9.3.2 Industrialization 189
9.3.3 Urbanization 190
9.3.4 Prestige 190
9.3.5 Use as school language 191
10 Language and national identity 193

10.1 Language and nationalism 193
10.1.1 Basic concepts 194
10.1.2 Language and nationalism 199
10.1.3 The political dimension of language 203
10.2 Language planning 204
10.2.1 Language development 205
10.2.2 Language determination and allocation of
language use 207
10.2.3 Language-planning policies - some examples 210
10.2.4 Practical considerations 214
11 Linguistic minorities 219

11.1 Indigenous minorities in western Europe 221
CONTENTS
11.2 Two main types of indigenous linguistic minorities 224

11.3 Ethnic, regional, national and minority languages 226
11.4 A historical perspective 227
11.5 The contemporary situation in western Europe 231
11.6 Use and maintenance of minority languages 233
11.7 Assimilation of minorities and cultural pluralism 235
11.8 Separatism and separate identity 237
11.9 Language and separateness 238
11.10 Attributes of minority languages 239
11.11 Non-indigenous minorities in Europe 242
11.12 The study of linguistic minorities 244
12 Case Study I: The Alsatians 248

12.1 The language 248
12.2 Historical overview 250
12.3 The use of Alsatian 254
12.4 The future of Alsatian 259
13 Case Study II: The Catalans 260

13.1 Introduction: Catalonia and the Catalans 261
13.2 Geographical and demographic overview 264
13.3 The Catalan language 265
13.3.1 A sociolinguistic historical overview 265
13.3.2 The impact of immigration 271
13.3.3 The present linguistic situation 272
13.4 Language planning since 1975 273
13.5 The public use of Catalan 277
13.6 The extent of Catalan knowledge 284
13.7 The future of Catalan 286
14 Case Study III: Migrant workers in the Federal Republic of

Germany 289
14.1 Old and new minorities 289
14.1.1 Indigenous minorities 289
14.1.2 Refugees and returning German minorities 291
14.1.3 New minorities 292
14.2 Migration 293
14.2.1 Background to migration to Germany 293
14.2.2 Composition of the new minorities and
geographical distribution 295
X . CONTENTS
14.2.3 Reasons for migration and problems

of remigration 297
14.2.4 Official policies towards migrants 298
14.3 Sociolinguistic aspects 300
14.3.1 Socio-demographic composition 300
14.3.2 The migrants’ linguistic background 301
14.3.3 Problems associated with migrants’
learning of German 302
14.3.4 Factors affecting migrants’ acquisition
of German 304
14.4 The education of migrant workers’ children 307
14.4.1 The EC Directive on the education of migrant
workers’ children 307
14.4.2 Educational provision for migrant workers’
children in the Federal Republic 308
14.4.3 Problems involved in the education of
migrants ’ children 310
Appendix: Council of the European Communities

Directive of 25 July 1977 on the education of the
children of migrant workers 314
References 316
Subject Index 339
Index of Places, Languages and Peoples 345
Author Index 350
List of maps
11.1 Indigenous linguistic minorities in western Europe 220

12.1 Alsace Lorraine 249
13.1 Linguistic minority areas in Spain 260
13.2 The spread of Catalan dialects 261
14.1 Migration and immigration to West Germany 290
Preface
During the last decade or so researchers into bilingualism have been

extraordinarily prolific, and considerable advances have been made in
our understanding of the subject. To those who have either a profes¬
sional or a personal interest (or both) in this area, it has indeed been
exciting to observe the progress made. At the same time, the task of
presenting an introductory study has been made more difficult, in that
it has become necessary to make more decisions about what aspects to
include in the discussion of each topic dealt with and which contribu¬
tions by scholars to exclude.
As an introduction to bilingualism, my aim in writing this book has
been to discuss the most important aspects and pertinent concepts of
individual and societal bilingualism, endeavouring to give some idea of
the complexity of some of the issues involved, as well as of the present
state of the relevant research. The need to be selective has led to the
decision to focus on the patterns of bilingual acquisition and use
among bilingual children in Part I and to the adoption of a European
bias in Part II, where examples are taken primarily from European
multilingual settings - the discussion of which necessarily reflects the
author’s own aquaintance with certain aspects of European culture.
Similarly, illustrations of linguistic points (e.g. of language use among
bilinguals) often involve the European languages of which the author
has either specialist knowledge or a degree of familiarity. This Euro¬
pean orientation may have the advantage of enabling the reader to see
how truly normal and widespread the phenomenon of bilingualism is,
even on a continent where most component countries still like to per¬
ceive themselves as monolingual nation states.
The structure of the book was originally conceived on the basis of
xii PREFACE
courses on Bilingualism taught to final-year Modem Language stu¬

dents and to postgraduates who had undergone training in European
languages. While no previous knowledge of the subject of bilingualism
is assumed here, I have presupposed familiarity with basic linguistic
concepts (e.g. ‘semantics’, ‘loan words’, ‘phonological development’).
Technical terms are given in inverted commas (e.g. ‘blend’, ‘code¬
switching’) or italics {language shift) where they may be first
introduced, but thereafter they may be used without inverted commas
or emphasis. In both Parts, the individual chapters have been organized
so that the general discussion precedes the examination of specific as¬
pects. It is possible, however, to read sections of the book on their own
or out of sequence; in this case, the cross-references can serve as poin¬
ters to related issues.
In preparing the pages that follow, I have drawn widely on my own
experience of bilingualism, which therefore needs to be explained
briefly. I grew up speaking two languages (Danish and German) and
was trained as a linguist; as a young adult, I moved from Germany to
Britain, where I settled and married a Spaniard. I use different lan¬
guages depending oa whom I am speaking to or what I am talking
about, and I feel a different kind of cultural and emotional attachment^-
to the languages involved. In other words, I use and relate to language
in the same way as many bi- or multilinguals, immigrants, migrants
and children of linguistically mixed parentage. The initial impetus for
pursuing the study of bilingualism was provided by my children, as the
experience of seeing bi- and trilingual language development at first¬
hand brought the whole subject alive. The same has happened to many
other linguist-parents, who have raised their children to use more than
one language and have recorded their observations, thus providing data
for the analysis of bilingual language acquisition. The study of child
language requires a good deal of close observation, for which parents
may be in a privileged position - one of which they are entitled to take
advantage of. This, I feel, explains the frequent mention of the lan¬
guage use of linguists’ own children (the present author’s included).
My experience of bilingual speakers of both sexes and all ages has
led me to adopt a usage of personal pronouns which may seem odd. I
employ ‘he’ or ‘she’, and also the less personal ‘(s)he’, in reference to
noun phrases such as ‘the bilingual’, ‘the individual speaker’ and ‘the
child’, because I often, though not always, had particular people in
mind.
A CAMPUS Scholarship awarded by my University enabled me to
V
PREFACE xiii
undertake research in the context of linguistic minority settings. Many

colleagues helped in the task of completing the book. My former
Chairman of Department, Martin Harris, provided both the initial
stimulus and subsequent patient support. Several members of staff in
the Department of Modem Languages at Salford gave me cheerful en¬
couragement, furnished me with examples, observations and anecdotes
about their own bilingual experiences, and generally commented on
parts of earlier drafts. I am particularly grateful to Andy Hollis for his
meticulous and good-humoured scmtiny of several chapters and for his
remarks on matters of content as well as style; without his suggestions
this book would contain many more Teutonisms and bilingual speech
markers than it does already (incidentally, I hope that he, too, now sees
himself as one of Europe’s bilinguals). Andrew Dobson, of Keele
University, kindly advised me on several points in Chapter 10, which
draws heavily on other disciplines. Alan Yates, of Sheffield University,
and my colleague Enlalia Torras, read Chapter 13 and made some use¬
ful comments. Suzanne Romaine read the final draft and made many
detailed and valuable points, thus generously sharing with me her
knowledge and understanding of bilingualism. To all these colleagues
my sincere thanks. Needless to say, I alone am responsible for any
shortcomings of content or infelicities of English style that remain.
Many friends and relatives in a number of European countries and
beyond demonstrated their support for my project in different ways
which I now gratefully acknowledge. My greatest debt is to my child¬
ren and my husband. Cristina and Pascual unwittingly provided
numerous instances of bilingual speech behaviour and, rather more
consciously, endured a good deal of discussion on the subject at the
family table. They rarely complained when my writing took preference
over family life. I thank them for their help and tolerance. My husband
was my ever-willing critic and sounding board; without his unfailing
support and his enthusiasm for bilinguahsm this book could not have
been written. I can only hope that it justifies the ilusion he imbued in
me.
The Publishers are grateful to the following for permission to repro¬

duce copyright material: The Society for the Psychological Study of
Social Issues, and the author, for a figure from Joshua A. Fishman,
‘Bilingualism with and without Diglossia; Diglossia with and without
Bilingualism’, Journal of Social Issues XXIII, No 2; Department of
XIV PREFACE
Sociology and Social Psychology, University of Tampere for a figure

from Skutnabb-Kangas, T. and Toukomaa, P. (1977) Teaching migrant
children’s mother tongue and learning the language of the host
country in the context of the socio-cultural situation of the migrant
family (Research Report 15); Multilingual Matters Ltd. for a figure
from Baker, C. (1988) Key Issues in Bilingualism and Bilingual Edu¬
cation, a table from Baetons Beardsmore (1982) Bilingualism: Basic
Principles and a table from Skutnabb-Kangas (1984) Bilingualism or
Not; A figure by Joan Rubin from Readings in the Sociology of Lan¬
guage (1968; 526), edited by J. Fishman. Copyright 1968 Mouton &
Co. Reprinted by permission of Mouton de Gruyter, a division of Wal¬
ter de Gruyter & Co.
NOTE: All references to the Federal Republic of Germany in this

book are to West Germany prior to German reunification in 1990.
This book is for
Paco, mi bilingiie favorito;
fiir meine Kinder Cristina und Pascual;
og til min mor, som startede det hele.

Introduction
Bilingualism - a natural phenomenon

To what extent can we ever speak, within a European context, of com¬
munities or languages which have developed in isolation, without
contact with other peoples, cultures and languages? One would have to
look to the outer fringes of the continent. Iceland and perhaps Portugal
are probably the only European states which are genuinely monoethnic
- but their inhabitants are not necessarily monolingual. The usual state
of affairs elsewhere in Europe, and in most of the rest of the world, is
that over the years and the centuries, communities establish links with
one another, whether friendly, commercial or belligerent. The experi¬
ence gained from such associations can be either positive or negative
for those concerned. One constant factor in such contacts, however, is
that they always influence in some way the speech behaviour of the
people involved - and therefore, eventually, their languages. Some¬
times the result is bilingualism, sometimes the languages are affected
by borrowing or change. With the passage of time, one community
may undergo gradual language shift, replacing bilingualism with
monolingualism, as is the case, for instance, in Brittany, where Breton
is giving way to French. Or a sudden change in the conditions of life
in a community may cause it to adopt a hitherto unused language on a
widespread basis, as became common during colonial times in many
parts of Africa and Asia.
Linguistic frontiers rarely coincide with political ones. The former,
however, if they are marked by geographical features, such as wide
waterways or high mountains, sometimes prove to be remarkably
stable, whatever happens to the latter. In general, both political and
linguistic demarcations are subject to change. Virtually everywhere
2 INTRODUCTION
countries have, in the course of their history, seen their borders expand
or retract according to political fortune. Yet historians seldom record
the linguistic impact on people made by occupation, annexation or en¬
forced change of allegiance as a result of war, political marriage or
some other cause. I find such scant reflection on linguistic matters
quite surprising considering the role language plays in our lives.
Centuries of political, economic and cultural interaction have made
Europe a continent with widespread bilingualism. At a political level,
most of the world’s sovereign states consider themselves to be rhbno-
linguairaTTatUtude held b>rdomTifa elites and nol!olibtlhfIuehce3T?y
Ae pervasiveness of political philosophies that promote the one-nation-
one-language ideal. But this ideal does not reflect reality as, at the
societal level, multilingualism exists, almost everywhere, as a result,
on the one hand, of historical events bringing about major shifts in
power and numerous border changes, and, on the other, of immigration
and migration.
As far as western Europe is concerned, it is necessary to add that,
whereas multilingualism at the level of society is common, bilingual¬
ism among individual speakers in some of the larger countries is less
frequent. There are a dozen or so languages spoken by sizeable com¬
munities in France and Germany, and well over a hundred in Britain,
but, in spite of this, the majority of Britons, Germans and French
peoplos.are monolingual in the sense that they use only one language
for their normal day-to-day communication.
Of the five languages spoken by 45 per cent of the world’s popula¬
tion (English, Spanish, Russian, Hindi and Chinese), the first three are
used in many countries as second or official languages, and Hindi
serves as a religious language for many, as well as being the national
language of India. There are an estimated 5000 tongues in the world
(Wardhaugh 1987; Crystal 1987), but only some 190 states, so it fol¬
lows that many countries must contain many different languages, i.e.
be multilingual. Approximately 95 per cent of the world’s population
are speakers of the 100 most frequently used languages.
In Europe only six countries are officially bilingual or multilingual:
Belgium, Finland, Greenland, Switzerland, the USSR and Yugoslavia,
and a handful of others accord official status of some kind to one or
more of the languages of their linguistic minorities, e.g. the Nether¬
lands to Frisian, Italy to German and the Federal Republic of Germany
to Danish. In western Europe people tend to live in monolingual envi¬
ronments where there are relatively few natural contact situations
THE ROLE OF ATTITUDES 3
requiring the use of more than one language, unless the speakers are
members of linguistic minorities. Being bom into a minority com¬
munity or into a bilingual family are the most common ways for
Europeans to become bilingual. There are, of course, other reasons -
personal, social or cultural - why individuals become bilingual. At one
time, for example, French was spoken by all members of the European
aristocracy as the use of this language signalled membership of the
elite. Prussia’s Frederick the Great once expressed the belief that Ger-
man was suitable only ‘for horses and soldiers’ and professed that he
spoke German ‘like a coachman’. Incidenfairy,” he also reft^^proofThat
he did not write it very well. Today, when Scandinavian, German and
Dutch technologists, academics or business people meet, they are like¬
ly to discuss their work in English, because this is the language most
widely used by such specialists for international communication. And
Remish-speaking Belgians embarking upon an administrative career in
Brussels need to be fluent in French. For many Europeans adding an¬
other language to their repertoire has become desirable and often
necessary. In other continents natural linguistic diversity is more pre¬
valent, and for many communities bilingualism is a normal
requirement for daily communication and not a sign of any particular
achievement.
The role of attitudes

Bilingualism arises as a result of contact. Whether it spreads
throughout the community and is maintained depends on whether the
conditions for its development are right. These, in turn, are determined
by individual and group attitudes towards (1) the two languages in¬
volved, and (2) bilingualism itself. In Europe there is considerable
linguistic diversity, yet this fact is rarely reflected in the official
policies adopted by most European countries. Few states encourage the
learning of minority languages, for instance, as part of the school cur¬
riculum for both minority and majority children. Attitudes towards
bilingualism range from officially promoting or accepting it, as is the
case in the officially bilingual countries mentioned earlier, to granting
some regions within a state the right to use the regional language in
administration and education, as happens in Spain with regard to Cata¬
lan, Basque and Galician. Government acceptance of bilingualism,
rather than active promotion, seems to be the attitude that underlies
bilingual arrangements in Wales, where some schools offer Welsh-me-
4 INTRODUCTION
dium instruction and others use both English and Welsh, while some
public institutions (but not all) offer services in both languages. Where¬
as West Germans and Danes may regard individual bilingualism
positively, especially if it involves languages which are considered
‘useful’ (e.g. English or French), they may express less admiration
about bilingualism of their Turkish migrant communities. In Britain,
the suggestion that the children of the majority might profit from learn¬
ing the languages of some of the immigrant groups (such as
Cantonese or Urdu) has been greeted by many people with indifference
and by some with outright hostility.
As a result both of political developments and social pressure, atti¬
tudes towards bilingualism are changing for the better in many parts of
the world. One of the landmarks in this process is what happened in
the United States during the 1960s. At the same time as the position of
minorities was being revised in the wake of the Civil Rights move¬
ment, the United States experienced large-scale immigration from
Spanish-speaking people who neither assimilated by Anglo-Saxon cul¬
ture and values nor shed their mother tongue as readily as previous
generations of immigrants had done. The debate that took place both in
intellectual and political circles eventually led to legislation supporting
bilingual education programmes.^ At about the same time Canada de¬
clared itself a bilingual state and officially enshrined the language
right^ of its French-speaking minority in law, by means of a number of
legislative measures. This came about after a series of political crises
prompted by the social and economic grievances of French Canadians
and the threat of secession by the province of Quebec. However, little
was done to promote the language needs of Canada’s many other mi¬
norities, such as the indigenous Inuit and the immigrants from Poland,
Ukraine and Hong Kong, to name but a few.
Bilingualism as a problem and as a resource

Individual bilingualism has, in the past, often been blamed for a biling¬
ual child’s underachi• afirTTr
has Tieen considered .toJlie.. at The root .gf-mijiarity m^bers’ lack of
assimilation to mainstream society. Some of these claims'^ analysed
in Chapters 6 and 7, where it is suggested that social causes, rather
than bilingualism, are usually responsible for problems of this kind.
Societal multilingualism can also be seen as a problem, if one be¬
lieves that language itself (not just people) can cause conflict. For
BILINGUALISM AS A PROBLEM AND AS A RESOURCE 5
example, Fasold (1984: 4) suggests that ‘multilingualism works against

nationalism’ in countries which are still undergoing a process of nation
building and where there is no universally agreed language to unify a
state’s diverse ethnic groups. Wardhaugh’s (1987) study of language
spread and decline focuses on issues raised by the competition of lan¬
guages within countries which have been linguistically invaded from
outside the national borders. He argues that some kind of conflict be¬
tween languages will ensue once the boundaries between them
(territorial, ethnic, political or functional) come under pressure, and
that this may result in a major change in a group’s language patterns;
When such boundaries are weak, the languages will not only be in contact,
they will also be in competition. As one gains territory, speakers, or func¬
tions, all others lose. Bilingualism may not be a real choice in such
circumstances; it may be no more than a temporary expedient, a somewhat
marginal phenomenon, because when one language encroaches on another,
bilingualism may prove to be only a temporary waystage to unilingualism
in the encroaching language as the latter assumes more and more functions
and is acquired as the sole language by greater numbers of speakers.
(Wardhaugh 1987: 17)
Any such process is likely to lead to tension and resentment on the

part of the group(s) involved. Speakers of the ‘losing language(s)’ may
find themselves pushed to the political, social or cultural margin, and
they may feel that their national or ethnic identity is under threat, along
with the survival of their language.
Language contact usually results in an uneven distribution of lan¬
guage patterns among the groups involved. If distribution were
identical, i.e. if the languages concerned were perceived as having
equal prestige and they were each used to the same degree and in equi¬
valent situations, then there would probably be no problem. But this is
not often the case.
A preoccupation with the problems associated with bilingualism
should not, however, obscure awareness of its likely benefits. In the
first place, individual bilingualism is often experienced as an enriching
attribute that facilitates a better understanding of the nature of language
and provides an opportunity for gaining a deeper insight into two cul¬
tures. However, this favourable perception appears to depend on the
socio-cultural context of bilingualism, as it is unlikely to be shared by
those members of linguistic minorities who find themselves discrimi¬
nated against by the dominant majority (which itself may have
6 INTRODUCTION
negative attitudes towards bilingualism). A second possible advantage

for members of multilingual groups is that they have at their disposal a
wider range of linguistic resources than those who belong to monolin¬
gual communities. This may enable them to communicate with others
in a more flexible and diverse way - a benefit which, in a world of
increasing international communication, should be obvious. The study
of bilingualism can provide greater awareness of the versatility of
social organizations and the resourcefulness of human behaviour. In
general, an appreciation of the issues involved in bilingualism may
help to eliminate prejudice or fear.
The remit of the study of bilingualism

An examination of bilingualism involves looking at the many factors
that contribute to its development, maintenance and loss, and at the
different ways in which individuals or communities respond to the lin¬
guistic challenges they encounter. Bilingualism and multilingualism are
not static conditions. In many immigrant groups the shift from the
‘old’ home language to that of the host country takes place over three
generations; but the process is sometimes more rapid, leaving only the
immigrant generation as bilingual. At other times, however, the shift
takes much longer, or does not occur at all, leading to well-established
bilingualism over many years. For the immigrant, learning to live in a
new culture or to accommodate to a new language is a dynamic pro¬
cess. Individuals go through phases in their lives when one of their
languages becomes weaker, or stronger, as the case may be; or they
add new languages to their daily communicative repertoire, developing
clearly defined situations in which each code will be used.
The field of study is multidisciplinary. We approach the study of
bilingualism from at least three different perspectives, i.e. those of:
(1) the individual;

(2) the group;
(3) the language systems.
A wide range of social sciences contribute towards our under¬

standing of bilingualism: sociology, ethnology, anthropology,
education, psychology and most branches of linguistics, including
pragmatics. Each participating discipline brings with it its own meth¬
ods and terminologies, and studies in bilingualism reflect the academic
background of the researcher. Language behaviour is highly complex.
THE REMIT OF THE STUDY OF BILINGUALISM 7
and no one academic subject alone can hope to explain it completely.

Linguistic performance is influenced by emotional factors, by the
speaker’s linguistic knowledge, and by perceived social values and
norms. The interplay of these (and other) determinants becomes even
more intricate if the speech patterns of whole groups of people are
under examination. Research into bilingualism must be multi- and
interdisciplinary so that all the relevant phenomena can be described
and accounted for. The ultimate aim, as in other social sciences, is to
establish a valid theoretical construct.
One of the problems encountered by researchers into societal bilin¬
gualism relates to the reliability of data. When considering patterns of
language use, one needs to handle such statistical information as is
made available by official (or national) agencies. But many countries
do not include questions on language use on their census question¬
naires, or they may ask only one question, without specifying function
or degree of competence, e.g. ‘What language(s) do you speak?’,
which can be interpreted in a variety of ways. Nelde (1984) warns that
language census data should always be approached critically, adding
that in areas where there is conflict and minorities are disadvantaged
results may not be at all meaningful. In such cases, he argues, the
replies of minority members tend to be influenced by a number of
extra-linguistic considerations which, in sociolinguistic analysis, it may
be difficult to come to terms with. For instance, census returns may
indicate a minority group’s aspired rather than actual linguistic affilia¬
tions, and this may favour the dominant rather than the minority
language. Socio-economic factors, too, may cause bias in answers to
questionnaires, usually in the direction of the majority language.
The systematic description of bilingualism, particularly using a
sociolinguistic approach, is of relatively recent origin. Einar Haugen’s
work on the Norwegian language in America (1953) and on bilingualism
in the Americas (1965) constituted the first large-scale sociolinguistic
analyses of bilingualism. Uriel Weinreich’s book Languages in Contact
(first published in 1953) quickly became a classic and was seminal in
developing research methods combining both social, psychological and
linguistic dimensions. Robert Lado’s Linguistics across Cultures (1957)
did much to arouse linguists’ interest in cross-cultural and bilingual
communication, and to open their eyes to the sociolinguistic reality of
language use, thus weaning them away from an exclusive preoccupa¬
tion with linguistic forms and norms.
Education plays a crucial role both in the personal and social devel-
8 INTRODUCTION
opment of the bilingual individual and for the success of societal multi¬
lingualism. It has been attested (e.g. by Skutnabb-Kangas 1984a), for
instance, that unless the educational system takes proper account of
minority children’s special language needs (in both languages), they
will not become fully functional in the minority and the majority
codes. On the other hand, a minority language that finds its way into
the school curriculum will enjoy enhanced prestige and this can, in
turn, positively affect public attitudes towards the language concerned
and its speakers, who may then find it easier to maintain it. The issue
of how best to educate minority children has become particularly ur¬
gent in many countries which have been affected by large-scale
migration and emigration. Research in the area of bilingual education
has often come as a response to such pressing needs.
Coverage and organization of this book

The organization of this book broadly reflects the two main strands in
the study of bilingualism: in Part I individual bilinguals and their lan¬
guage use constitute the central focus, while in Part II a broader
sociological and sociolinguistic approach is adopted for the discussion
of multilingualism in society. Educational matters are raised in both
parts, as they affect the development and maintenance of bilingualism
in both the individual and society.
Part I seeks, from a microlinguistic point of view, to embrace both
the psycholinguistic and the purely linguistic dimensions of bilingual¬
ism. Starting with the discussion of some definitions of bilingualism.
Chapter 1 looks at individual bilingualism in a general way. Chapter 2
narrows the discussion to bilingual children, while Chapter 3 focuses
specifically on the bilingual acquisition process. Chapters 4 and 5
examine the linguistic perspective, concentrating on bilingual com¬
petence and speech on their own terms, i.e. from the bilingual (rather
than the monolingual) angle. In the last two chapters (6 and 7) of this
part the discussion is broadened again to take in some of the major
factors of a social, cultural, psychological and cognitive nature that are
considered to be important in the acquisition and maintenance of bi¬
lingualism in the individual. In line with the general bias of this book,
in Chapters 3, 4 and 5 certain aspects of bilingualism in children are
given prominence; on the other hand. Chapters 5 and 7 deal with issues
which apply to both bilingual children and adult users of two or more
languages.
COVERAGE AND ORGANIZATION OE THIS BOOK 9
1- ,
In the macrolinguistic approach adopted in Part II, the discussion

ranges from the general to the more specific. Chapter 8 looks at the
different patterns of multilingualism which can result from language
contact and at some of the factors which are influential in the emer¬
gence of societal multilingualism. Questions such as ‘By what criteria
do bilingual communities organize their language behaviour?’ and
‘Why do some of them seem to be so much more successful than
others in maintaining their language(s)?’ are dealt with in Chapter 9.
Decisions by the state on whether to promote or suppress particular
languages can have a powerful effect on people’s language behaviour,
and the aims of Chapter 10 are to assess the possible role of language
for social and national identification, and to look at how decisions on
language matters are made by the state. Chapter 11 touches on a wide
area, as the discussion of linguistic minorities must take into account
many of the major issues in sociolinguistics: contact between dominant
group and minority; the linguistic behaviour of both communities; atti¬
tudes towards the groups and languages involved; questions of identity,
conflict and integration; and topics relating to language provision,
maintenance and shift. In order to illustrate the complex issues raised
in Chapter 11, three linguistic minorities that have little in common
with each other, and that have perhaps been less accessible to English-
language readers, have been chosen for more detailed consideration.
The aim here is to show that language contact in Alsace, Catalonia and
West Germany manifests various degrees of intensity and distinct dy¬
namics, and that this results in different patterns of bilingualism.
The European dimension of this book has made it possible, on the
one hand, to be selective in the choice of examples. On the other hand,
choosing Europe as a general frame of reference has entailed the ex¬
clusion of certain important aspects of the study of bilingualism,
pidgins and creoles, for instance. Neurolinguistic aspects of bilingual¬
ism, issues of language contact and language change and the question
of attitudes towards bilingualism can be seen, perhaps, as less central
for an initial introduction to the subject and have therefore not been
included. The interested reader should turn to other introductory works
(e.g. Appel and Muysken 1987, Hamers and Blanc 1989, and Romaine
1989).
In many books and articles the words multilingualism and, occa¬
sionally, plurilingualism are used almost interchangeably with
bilingualism, the difference being quantitative rather than qualitative.
The former two are favoured in some of the European linguistic tradi-
10 INTRODUCTION
lions (e.g. in Belgium, and also in Germany and Switzerland, where

the subject is more frequently referred to as Mehrsprachigkeit (‘multi¬
lingualism’), than Zweisprachigkeit (‘bilingualism’). The use of the
term multilingualism allows one to take a broader view of language
and dialect varieties. There are even scholars (e.g. Wandruszka 1979)
who speak of ‘intralingual multilingualism’, referring to an ability to
operate on a number of different levels of style, register and dialect in
the mother tongue.
Hamers and Blanc (1989) use the terms ‘bilinguality’ and ‘bilingual¬
ism’ as reflecting a different level of analysis. The former, for them,
refers to the ‘psychological state of the individual who has access to
more than one linguistic code as a means of social communication’ (p.
265). Bilingualism, on the other hand, is defined in a more general way
as ‘the state of the individual or a community characterized by the
simultaneous presence of two languages’ (ibid.). This distinction may
have its advantages for certain types of analysis, but it has not yet
found wide acceptance and it is not adopted here. Normally, bilingual¬
ism is used in the literature for individuals and communities in which
two languages are present; the term multilingualism can refer to so¬
cieties where more than two languages are found. They are essentially
the same phenomenon. Both terms are employed in this book in the
description of language contact at the societal level, but only the for¬
mer is made use of in the discussion of individual bilingualism. When
speaking of a bilingual or multilingual community we are referring to
the presence of two or more languages in that particular setting, with¬
out implying that all (or most) of the members of the group in question
have competence in those languages. On the other hand, when adopt¬
ing a linguistic or psycholinguistic approach in the study of the
bilingual individual, the use of ‘bilingualism’ does imply the ability to
use both codes.
Notes
1. It should be noted, however, that in the 1980s the commitment to
bilingual education shifted notably. The issue was no longer felt
to be one of national federal concern, and public funds were
withdrawn from many schemes; it was left to individual states to
decide the extent of official support for bilingualism.
PARTI
PSYCHOLINGUISTIC ASPECTS OF
BILINGUALISM
u*-
irtfW’V kL^ «»T ■R#iiy‘4.11

'4l>i fstJe^.l !•> "’(:? ■Ô-l !.’»' yt «• ■ ■-
'36 aToaîsA 'oiT3iuSMiJ6ffQy^3

M8IJAU0V1tjlta
r ’4r)s- ^$^..1^
(w' ..>.^
<rf wyl^v f-:;
(7î5u^ j“c ofcm^ '^biftf»,.’aal«< sod 't^rn^ffoS

kfimr, «'4f ih«T).
* i*Ei® bu
tD
fjA* eotli: M e^ amiaanmiiun:’ 4*1

T!fi tissc; -31^5/ ?8>it ;i«!|^ »i i
j.% 'rtif St
'pi^\\ ^znpi^',‘ \êi.X T5ib cfcJtiiiCLv'>i ovtsy
e >1.*^ OU >«!_
" )U'- $s 1«|!ic4<1)I' - l£?' r-#' ■•'SMki irdi ?'.^?T8tttuwiiii»

,s*V.,W4 reft*./' 1^3 ^ • i
••>.0 . *9^
^ J9'^.=Ji> .eo» ifi'W
^ a »lii- yK**'A4 San ^?^l^ ^ iSjf-
***;. X f^|g^1tt[tic^'^'ijft/u*e iiMtv-»ici4jî^ v&^^tajjauv Wifcij,/
_‘*^ WfB«? xfrnîn
S*^-& ’^r !;r .^«( |4»f»t'ukP icsliflsi. ♦i»fi-
^ (..•) ;.V:,y •
‘ , 4 ')i_ «1^ -O£r*';r'0-Jp^ ’*itm

kv. :<xJ» .'n'i«Jy ol;
rtP^v^«I. Uit of iJl' »
Uhf bc^ iAi^
V*t.:
Notci
1, i* rvasiKj, >|ovvtv«ri &J ttii? |930» i>»e ««^rim»j}ft«i|ii ii»j
!fci,|fe?;.^.>0’'y)t .••-■ ♦ 4» fck
'4y f^,tem^ ct>ef-Tiv, mvA p«|44^ fkiWi* s»v^
‘ tt'fV'.drrfwii ft >04,1 r.trt>i*«;s: a H'w toik'Sh^Wv*! jiU(*r<? m
ik Jl*- <»tef!K '.<#■ i *1 :;,<■-*■
. 'k^'.
Chapter 1
Individual bilingualism
1.1 Societal and individual bilingualism
When looking through introductions to the study of bilingualism one

sees that it is often mentioned that over half the population of the
world is bilingual. Such general statements are open to different inter¬
pretations, in view of the fact that patterns of language use found in
bilingual communities can be quite varied and that many are changing
even as we observe them.
Statistics can mislead, particularly when they do not distinguish be¬
tween societies and individuals, as there is a fundamental difference
between societal and individual bilingualism. We may say that India,
Switzerland and Belgium are multilingual countries, that Canada is of¬
ficially bilingual, as is Finland, or that Paraguay is an example of
bilingualism and diglossia, and Luxembourg one of triglossia. In such
contexts the labels ‘bilingual’ and ‘multilingual’ reflect official policies
towards some, or all, of the countries’ minorities. On the whole, how¬
ever, they say nothing about the degree or the extent of bilingualism
among the inhabitants of these areas. It is only when language plan¬
ning policies find their way into a nation’s education system with the
explicit aim of fostering bilingualism (rather than promoting the ma¬
jority or the minority language) that bilingualism may be the norm in
such parts of the world. Of the countries mentioned above, individual
bilingualism may be widespread only in Luxembourg and Paraguay.
Mackey (1970) claims that there are actually fewer bilingual people in
bilingual countries than there are in so-called unilingual ones, and he
points out that the main concern of multilingual states has often been
the guaranteed maintenance and use of two or more languages in the
14 INDIVIDUAL BILINGUALISM
same nation, rather than the promotion of bilingualism among its

citizens. This observation is particularly relevant in the non-European
context, since in western Europe we find only three states which are
officially multilingual, Belgium, Switzerland and Finland, and the first
two pursue policies of territorial monolingualism (see Chapter 8).
In this part of the book attention focuses on the bilingual speaker as
the ‘locus of language contact’ (Weinreich 1968: 1). Naturally, lan¬
guage interaction involves communication in a wider sociocultural
setting. We therefore need to bear in mind certain aspects of the con¬
text in which the bilingual finds herself or himself, as this provides us
with information about the ways in which the two or more languages
form part of the person’s everyday life. Equally, it may be relevant to
take note of psychological factors that may influence, or affect in some
way, the linguistic behaviour of the bilingual, just as one must also be
aware that neurological, pathological and general cognitive factors can
come into play.
1.2 Describing bilingualism
The most salient feature of bilingualism is that it is a multi-faceted

phenomenon. Whether one is considering it at a societal or an individ¬
ual level, one has to accept that there can be no clear cut-off points. As
bilingualism defies delimitation, it is open to a variety of descriptions,
interpretations and definitions. We can consider some examples. In
Britain people do not usually think of Wales as a bilingual part of the
state, yet one does happily use the label ‘bilingual’ when referring to
certain types of schools found in the principality. Many people would
readily call ‘bilingual’ the two-year-old child of a French-English
couple, and the fact that the toddler’s vocabulary may consist of some
200 French and English items in all does not seem to be of importance.
Similarly, size of vocabulary may not carry much weight in the case of
a graduate in French, who may have spent a considerable amount of
time in France and studying the language, and whose total lexicon will
be several hundred times larger than the child’s. This person is not,
however, often thought of as a ‘bilingual’, and would not normally
claim the label for himself or herself. ‘Multi-cultural’ and ‘multi-eth¬
nic’ are adjectives freely used by many people in the English-speaking
world, and the children who start school in the UK with little knowl¬
edge of English may be referred to as ‘minority children’ or ‘ESL
DESCRIBING BILINGUALISM 15
(English as a second language) pupils’, but not as ‘bilinguals’. Why

should all this be so?
1.2.1 Some deflnitions of bilingualism

The notion of bilingualism is firmly established in the mind of the lay
person. It may be tinged with bias, and it frequently carries either posi¬
tive or negative connotations. In the specialist’s mind the concept is
also well established. However, the latter is expected to apply objective
criteria and to aim for precise delineations. Yet some of the definitions
of bilingualism that have been put forward are surprisingly vague, and
even contradictory.
Uriel Weinreich, one of the founding fathers of bilingual studies and
a bilingual himself, offers one of the shortest definitions in his well-
known book Languages in Contact: ‘The practice of alternately using
two languages will be called bilingualism, and the person involved,
bilingual.’ (Weinreich 1968: 1).
An oft-quoted definition is found in one of the early books on mod¬
em linguistics, Leonard Bloomfield’s Language, first published in the
USA in 1933. When mentioning that foreign language learning among
immigrants may result in language shift, Bloomfield pays special atten¬
tion to users who become so proficient in the new language that they
are indistinguishable from the native speakers around them. He says:
In the cases where this perfect foreign-language learning is not accompa¬
nied by loss of the native language, it results in ‘bilingualism’, native-like
control of two languages. After early childhood few people have enough
muscular and nervous freedom or enough opportunity and leisure to reach
perfection in a foreign language; yet bilingualism of this kind is commoner
than one might suppose, both in cases like those of our immigrants and as
a result of travel, foreign study, or similar association. Of course, one can¬
not define a degree of perfection at which a good foreign speaker becomes
a bilingual: the distinction is relative.
(Bloomfield 1933: 55-6)
No doubt Bloomfield had a clear notion of bilingualism, but his defini¬

tion and subsequent qualifying remarks are not without some degree of
contradiction: if one cannot define ‘a degree of perfection’ in bilingual¬
ism, how can we talk of ‘perfect foreign-language learning’?
In his article ‘The description of bilingualism’, William Mackey of¬
fers a definition that incorporates Weinreich’s alternate use of two
languages and is preceded by Bloomfield’s reservation with respect to
the degree of proficiency:
It seems obvious that if we are to study the phenomenon of bilingualism

we are forced to consider it as something entirely relative. We must more¬
over include the use not only of two languages, but of any number of
languages. We shall therefore consider bilingualism as the alternate use of
two or more languages by the same individual.
(Mackey 1970: 555)
1.2.2 Some types of bilinguals

The three definitions mentioned so far say nothing about how well the
languages need to be known, whether both have to be mastered in all
sorts of skills, whether they must be used in similar or different situ¬
ations, or about any particular requirements regarding the uses to
which the languages are put. Yet such considerations would probably
be relevant in deciding whether any, or all, of the following should be
considered as bilinguals:
(1) the two-year-old who is beginning to talk, speaking English to

one parent and Welsh to the other;
(2) the four-year-old whose home language is Bengali and who has
been attending an English playgroup for some time;
(3) the schoolchild from an Italian immigrant family living in the
United States who increasingly uses English both at home and
^ outside but whose older relatives address him in Italian only;
(4) the Canadian child from Montreal who comes from an English-
speaking background and attends an immersion programme
which consists of virtually all school subjects being taught
through the medium of Erench;
(5) the young graduate who has studied French for eleven years;
(6) the sixty-year-old scholar who has spent a considerable part of

her life working with manuscripts and documents written in
Latin;
(7) the technical translator;
(8) the personal interpreter of an important public figure;
(9) the Portuguese chemist who can read specialist literature in his
subject written in English;
(10) the Japanese airline pilot who uses English for most of his pro¬
fessional communication;
(11) the Turkish immigrant worker in the Federal Republic of Ger¬

many who speaks Turkish at home and with his friends and work
colleagues, but who can communicate in German, in both the
written and the oral forms, with his superiors and the authorities;
(12) the wife of the latter, who is able to get by in spoken German but
cannot read or write it;
(13) the Danish immigrant in New Zealand who has had no contact
with Danish for the last forty years;
(14) the Belgian government employee who lives in bilingual Brus¬

sels, whose friends and relatives are mainly Flemish speakers but
who works in an entirely French-speaking environment and
whose colleagues in the office (whether they are Flemish or not)
use French as well;
(15) the fervent Catalanist who at home and at work uses Catalan
only, but who is exposed to Castilian Spanish from the media
and in the street and has no linguistic difficulty in the latter lan¬
guage.
So what is bilingualism? Many spiecialists would say that all the

above individuals could be classed as bilinguals; but public opinion,
and at least some of these people themselves, would probably disagree.
It is possible to think of a number of explanations for the difficulties
involved in arriving at a precise definition. The elusiveness of the phe¬
nomenon has already been referred to. Another factor is the nature of
the subject itself. Language use is part of human behaviour, and as
such not readily accessible to scientific investigation and experimental
research. The study of bilingualism is hampered by a host of methodo¬
logical problems and theoretical shortcomings. These difficulties stem
from the complex interplay and variability of social, psychological and
chance factors which determine individual conduct - and therefore
often render generalizations invalid. Research into bilingualism is, con¬
sequently, interdisciplinary in character, as scholars from different
academic fields, such as sociology, psychology, linguistics, anthropo¬
logy and education (and others) bring different methods, criteria and
assumptions to bear upon studies of bilingual situations.
‘Bilingualism as a concept has open-ended semantics’, Hugo

Baetens Beardsmore (1982: 1) points out at the beginning of his com¬
prehensive discussion of a number of different definitions of
bilingualism. Ultimately, all definitions are arbitrary to a greater or
lesser extent. It is not necessarily a problem, therefore, that there are so
many of them, since in this way the researcher is able to choose the
one that best suits her or his purpose.
1.2.3 Factors taken into account when describing bilingualism

One of Baetens Beardsmore’s contributions to the subject under dis¬
cussion is his exposition of a series of descriptive labels, which have
been chosen by various specialists so as to provide a frame of refer¬
ence. There are many and different areas around which the study of
bilingualism can be centred; some of them are fairly clear-cut, whereas
others are not. The following list contains some selected examples (a
fuller discussion of the issues involved can be found, for instance, in
Baetens Beardsmore 1982 or in Skutnabb-Kangas 1984a).
(1) The age of the bilingual at the time of the acquisition may result in
considerable differences, as suggested by the terms ‘early bilingualism’
and ‘late bilingualism’. An early bilingual may be a case of ‘infant
bilingualism’ (Haugen 1956: 72) or of ‘child bilingualism’. The cut-off
point is not firmly established, but it can be set arbitrarily (see, for
example, McLaughlin 1984: 73) at the age of three - and between the
child bilingual and the case of ‘adult bilingualism’ at the age of pu¬
berty. Tove Skutnabb-Kangas’s (1984a: 80 ff.) analysis of definitions
of bilingualism recognizes four main types. Her first type comes under
the heading of ‘origin’, which at this stage can be taken to correspond
with ‘age’. This factor, however, is more useful for the description
than for a definition of bilingualism.
(2) We can widen the ambit of the discussion to include considerations

of context. The infant/child who acquires two languages from the speak¬
ers around him/her in an unstructured way can be called a ‘natural
bilingual’ or a ‘primary bilingual’ (Houston 1972). Adler (1977) refers
to this specification with the expression ‘ascribed bilingualism’. The
two languages may have been presented to the infant/child bilingual
either in a ‘fused’ context (both parents using both languages to the
child; it can also mean that the context of acquisition is such that the
two languages are spoken in the same locality where the young learner
is beginning to use them - as for example in a multilingual society) or

in ‘separate’ contexts (the parents follow the one-parent-one-language
principle, or one language is learnt in one country and the other in a
different one). The person who becomes bilingual through systematic
or structured instruction (that is, undergoing some kind of training) is
also known as a ‘secondary bilingual’, and the result can be called
‘achieved bilingualism’ (Adler 1977). Skutnabb-Kangas (1984a: 95 ff.)
establishes the dichotomy of ‘natural bilingualism’ and ‘school bilin¬
gualism/cultural bilingualism’. The distinction between these two is
interesting. School bilingualism is, as its name implies, involved with
formal language teaching at school, during which the learner (as the
author points out) does not normally have much opportunity to practise
the language outside the classroom environment. Cultural bilingualism
coincides in many respects with school bilingualism, but it is more
often the result of language learning by adults who wish to use a
foreign language for purposes of travel, leisure or work; the assumed
cultural value (that an educated person is one who knows one or more
foreign languages) is reflected in the term.
(3) The relationship between sign and meaning, i.e. the mental organ¬
ization of the speech of bilinguals, was first mentioned by Weinreich
(1968), whose pioneering work was very much concerned with the
phenomenon of interference, that is, the influence of the bilingual’s
language systems upon each other. He distinguishes different types of
bilinguals according to the relationship that exists between the linguis¬
tic sign and the semantic content. In Type A the individual combines a
sign (‘signifier’ is the term used by Weinreich) from each language
with a separate unit of content (or ‘signified’, or ‘semanteme’). In
Type B the subject identifies the two signs (‘signifiers’) but regards
them as a single compound, or composite, unit of meaning (‘signi¬
fied’). His examples, using English and Russian, are:
(A) ‘book’ ‘kniga’ (B) book’ =‘kniga’
/buk/ /kn’iga/ /buk/ /kn’iga/

Weinreich also considers a third possible interpretation of the sign.

Type C relates to people who learn a new language with the help of
another (Weinreich calls this process the ‘indirect method’), i.e. by
finding equivalent signs (signifiers, or words). For example, ‘to an
English learner of Russian, the signified [or referent, i.e. the object that
the word refers to] of the form /kn’iga/ may at first be not the object
[the thing itself, the book], but the English word book, with which he
is already familiar’;
‘book’
/buk/
/kn’iga/
(Weinreich 1968: 9-10)
Weinreich calls his Type C ‘subordinative’ bilingualism, as it de¬

scribes the kind of bilingual whose second language or L2 is seen to be
very much influenced by his/her first language or LI (see also Paradis
1977b), whereas the expression he uses for Type A is ‘coordinative’. In
later research by Ervin and Osgood (1954) Type A is called ‘coordi-
nate’”- bilingualism and Types B and C considered together as
‘compound bilingualism’. This latter notion, in particular, has been in¬
terpreted differently by a number of linguists, and a variety of
hypotheses have been formulated, in relation with contexts, linguistic
ability and age at the time of acquisition (see, for instance, Lambert et
al. 1959; Macnamara 1967b; Jakobovits 1968; Lambert 1972; Baetens
Beardsmore 1974 and Paradis 1977a; for a detailed discussion of the
question, see Skutnabb-Kangas 1984a; 98 ff.). Others have cast some
doubt on the validity of such an attempt at classification (Diller 1972)
or claimed that there can be little significance in the distinction, even if
all the types could be convincingly established (Albert and Obler
1978). Wolk (1984: 125) points out that the processes of mental trans¬
fer and linguistic behaviour tend to be affected significantly by a series
of sociolinguistic factors, and that this interplay has largely been disre¬
garded in previous research, thus fudging the basis for the description
of bilingualism. He claims that distinguishing carefully among the vari¬
ous factors and processes at play will make it easier to describe
bilingualism on the basis of clear criteria: ‘Vor allem die Unterschei-

dung zwischen soziolinguistischen Bestimmungfaktoren und
kognitiven Vermittlungs - oder sprachlichen Verhaltensprozessen kann
bessere Ansatze und Kriterien fur die Beschreibung und das Verstand-
nis der Zweisprachigkeit liefern. ’
(4) The order and consequence of bilingual language acquisition is
reflected in such labels as ‘incipient bilingualism’ (Diebold 1961) and
‘ascendant bilingualism’ (Baetens Beardsmore 1982), which indicate
an increase in the person’s ability to use two languages, whereas ‘re¬
cessive bilingualism’ (Baetens Beardsmore 1982) points to a decrease.
Qualitative, as well as quantitative, judgement is expressed by the
terms ‘additive bilingualism’ and ‘subtractive bilingualism’ (Lambert
1974). The former implies that the addition of a second language to a
person’s first can result in enriched, or at least complementary, social,
cognitive and linguistic abilities, whereas the latter suggests that the L2
is learnt at the expense of the LI. As a consequence of various social
pressures, many minority groups in Europe find themselves undergoing
a process of language shift, away from their ethnic tongue and towards
the national language of the country they now live in. This means that,
although they are becoming more proficient in the L2, they are losing
skills in the LI; therefore, as the latter is not being maintained, it is
actually being ‘subtracted’ from their bilingual proficiency.
(5) One of the most challenging aspects to address concerns the ques¬
tion of how proficient a person needs to be in both languages. There
are a number of definitions based on this criterion of competence.
Some authors, as was seen earlier, define bilingualism as ‘near-native
control of two or more languages’ (Bloomfield 1933: 56), or ‘complete
mastery of two different languages without interference’ (Oestreicher
1974: 9), or see the bilingual as ‘a person who knows two languages
with approximately the same degree of perfection as unilingual spea¬
kers of those languages’ (Christopherson 1948: 4). Such definitions
express a perfectionist or maximalist view. The labels used are, for
example, ‘perfect bilingualism’, ‘true bilingualism’ and ‘ambilingual-
ism’. Halliday, McIntosh and Strevens (1970: 141) describe an
‘ambilingual’ as a speaker who has complete control of two languages
and makes use of both in all uses to which he puts either. True ambil¬
ingual speakers are very rare creatures. Who ever has identical
linguistic input and output in both languages? And who would habitu¬
ally use both languages for the same purposes, in the same contexts?
At the other end of the pole we can place those definitions which
express a minimalist stance. Haugen (1953: 7) sees ‘the point where a
speaker can first produce complete meaningful utterances in the other
language’ as the beginning of bilingualism. Others, such as John Mac-
namara, see a minimal degree of competence, in one of the four
language skills (speaking, writing, reading and understanding speech)
as sufficient: ‘I shall consider as bilingual a person who, for example,
is an educated native speaker of English and who can also read a little
French. This means that bilingualism is being treated as a continuum,
or rather a series of continua, which vary among individuals along a
variety of dimensions’ (Macnamara 1969: 82).
Where one view is obviously too narrow, the other is too broad to
be of much help. However, somewhere in the middle of our continuum
we can accommodate the notion of ‘equilingualism’ or ‘balanced bi¬
lingualism’. We would expect a balanced bilingual to possess roughly
equal proficiency of the two languages, but with no implication that
the knowledge this bilingual has in either language is compared to
monolingual standards. The term ‘balanced bilingual’, as used by Lam¬
bert, Havelka and Gardner (1959), is intended to refer to individuals
who are fully competent in both codes. As in the case of the ‘ambilin-
gual’, the ‘balanced bilingual’ is likely to be something of an ideal,
since most bilinguals tend to be more fluent or generally proficient in
one ),anguage, or at any rate in some uses of it, i.e. they will have a
stronger or ‘dominant’ language and a weaker one. (There is, inciden¬
tally, a convention among scholars to list the dominant language first,
so that a Spanish-German bilingual should not be confused with a
German-Spanish one.)
The language a bilingual feels more at home in, the ‘preferred lan¬
guage’, may coincide with the dominant one, but this will not
necessarily happen in every case. The experienced immigrant techni¬
cian who has made good progress in her profession in her country of
adoption, but who still has strong emotional ties with the people and
the culture of her country of origin, could constitute a valid example. It
is also possible for the dominant and the preferred language to change
roles in the course of the bilingual’s life.
The notion of relativism, as first expressed by Bloomfield (see his
definition earlier), is a central one in the discussion of any type of
bilingualism. Most bilingualism can be identified only in relative
terms. If we accept that there are degrees of bilingual competence, this
implies that bilingualism is measurable. The question of measurabilty
has received a good deal of attention (see Kelly 1969; Baetens Beards-
more 1982; Skutnabb-Kangas 1984a; Saunders 1982a), and useful
insights can be obtained from the descriptions of tests, as well as the
discussion of relevant criteria, for appraising bilingualism. The fun¬
damental problem lies in the question of norms. What (or who)»should
be the basis for the assessment? Monoglot terms of reference are often
used, but they should not be, as they can be relevant only in the (ex¬
ceptional) case of an ambilingual whose control of the two language
systems may be almost ‘complete’ (assuming we know what this
means) and who displays no traces of interference. For the vast ma¬
jority of bilinguals, ‘bilingual competence’ is not measurable in terms
of monolingual standards. It is, of course, possible to provide a de¬
scription of a bilingual person’s proficiency in Language A and in
Language B using accepted methods of language testing. But over and
above these two competences there is the bilingual’s specific com¬
petence, which may manifest itself in such features as switching from
one code to another (code-switching) or making use of items or lin¬
guistic knowledge from one language when speaking the other
(mixing), which may enhance his/her communicative ability (see Chap¬
ter 13).
(6) Statements about most cases of bilingualism centring on the de¬

scription of the relative proficiency in each language usually fall short
of including references to the use or function that the bilingual s lan¬
guages fulfil. Function-based definitions reflect the view that language
is not an abstract entity, but a tool employed for taking part in acts of
communication. Thus, for example, the definitions by Mackey and
Weinreich (see above) point to the bilingual’s habit of alternating be¬
tween the two languages. Els Oksaar (1966 and 1983: 19) suggests a
combination of the criteria of competence and function when she
defines bilingualism as: ‘the ability of a person to use here and now
two or more languages as a means of communication in most situations
and to switch from one language to the other if necessary.’ To this she
adds that such a broad definition needs to be made more specific ‘ac¬
cording to the languages and the situation. Reality does not demand of
a multilingual that he use his languages in all kinds of situations.’
Skutnabb-Kangas (1984a) explains the emergence of function-re¬
lated definitions in the context of the development of linguistic thought
in general and ideas about bilingualism in particular. The very narrow
delimitation of language competence put forward by some theoretical
linguists of the transformational-generative school, and especially the

notion of the ‘ideal speaker/hearer’ (i.e. ‘someone living in a complete¬
ly homogeneous speech community’, according to Chomsky) could not
satisfy those whose interests moved away from the descriptions of
grammars and towards the variability of language uses and language
users. Mackey introduces an article on bilingualism with the categori¬
cal statement that ‘Bilingualism is not a phenomenon of language; it is
a characteristic of its use. It is not a feature of the code but of the
message. It does not belong to the domain of langue but of parole'
(Mackey 1970: 554).
Just as we have to accept linguistic proficiency as something vari¬
able and unstable, we must acknowledge the existence of varying
degrees of functional bilingualism. Different bilinguals have distinct
uses, as well as various levels of competence, for each code. Where
precisely the threshold of bilingualism is set will depend on whether
one takes a maximalist or a minimalist viewpoint, and also on the fac¬
tors that one considers relevant for the description. For instance, we
may talk about ‘receptive bilingualism’ in relation to a person who is
able to understand a second language, either in its spoken or written
form, such as the Turkish wife in Germany mentioned before or the
Portuguese chemist. The term ‘passive’ bilingualism covers more or
less the same ground, although it has certain negative connotations and
does not do justice to the active decoding processes involved in under¬
standing language.
At the other end of the scale, functional bilingualism can be de¬
scribed in terms of ‘productive (= active) bilingualism’, which implies
the ability to speak as well as understand the languages, to write as
well as read them. There are many intermediate stages which can be
illustrated with examples from language learners who follow different
types of programmes designed to improve their language skills, both
oral and written, or cases like that of the immigrant who enters his new
country of residence with only a limited knowledge of either the
spoken or, perhaps, the written language, and who over the years adds
to his active and passive ability in the new code while at the same time
maintaining his mother tongue. There are, clearly, many conceivable
patterns of individual bilingualism, as can be seen by looking at the
table in which Baetens Beardsmore represents some of the possible
combinations for bilingual ability across the four language skills (al¬
though no information is given on the extent of skill for each
language):
TABLE 1.1 Patterns of individual bilingualism
Language skills Productive bilingualism
Listening comprehension: LI L2 LI L2 LI L2 LI L2
Reading comprehension: LI L2 LI L2 LI L2
Oral production: LI L2 LI L2 LI L2 LI L2
Written production: LI L2 LI
Language skills Receptive bilingualism
Listening comprehension: LI LI L2 LI LI L2 LI
Reading comprehension: LI LI LI L2 L2
Oral production: LI LI LI
Written production: LI LI LI
Source: Baetens Beardsmore 1982: 17
It is not difficult to find examples of the various combinations sug¬

gested above, particularly if one bears in mind that in many parts of
the world biliteracy is not to be assumed as an attribute of active bil¬
ingualism - just as bilingualism among young children usually
excludes any form of literacy. The last case, i.e. that of the receptive
bilingual who is illiterate in his LI but is able to read in the L2 (as it
emerges from the end of the last column in this table) is perhaps un¬
common in western Europe today. However, one can come across
people (e.g. Muslim immigrants) who cannot read in the language they
speak habitually but have learnt to read the Holy Scriptures of their
religion, i.e. the Koran, in Arabic. Another example would be the adult
immigrant who came to Britain from the sub-Asian continent and at¬
tended adult literacy programmes; he may have learnt to read English
to some extent yet have difficulties in following many types of conver¬
sation in this language; and he may be unable to write English at all.
(7) Skutnabb-Kangas (1984a) includes the issue of attitude in her dis¬

cussion, a point taken up by Kielhdfer and Jonekeit (1983: 11), who
stress also that it is crucial that the individual should be conscious of
his bilingualism. What is meant by ‘attitude’ is self-identification or

identification by others, i.e. whether a bilingual feels herself or himself
totally ‘at home’ in, or identified with, his or her languages, and
whether the individual is accepted by others as belonging to both the
one and the other linguistic community. Attitudes are more accessible
to observation in the context of societal multilingualism, as for
example in the case of bilinguals among minority groups, where it is
easier to notice that cultural, social and motivational factors can in¬
fluence a group’s maintenance or loss of bilingualism. The child whose
parents speak different languages, or whose home language is different
from that of the society around him or her, may well be bilingual but
not have developed strong links with other speakers of the parents’
original speech community, or with the culture associated with the lan¬
guage. In the case of many families, it can prove very difficult to
achieve such sociocultural embedding in both languages/ cultures, as
this requires a conscious effort on the part of the parents and also a fair
amount of access to the other country and its culture.
Some of the definitions presented can only be applied to specific
purposes - which may, of course, be exactly what was intended. Other
authors try to combine two or more criteria. Skutnabb-Kangas goes
further when she says:
A bilingual speaker is someone who is able to function in two (or more)

languages, either in monolingual or bilingual communities, in accordance
with the sociocultural demands made of an individual’s communicative and
cognitive competence by these communities or by the individual herself, at
the same level as native speakers, and who is able positively to identify
with both (or all) language groups (and cultures) or parts of them.
Source: Skumabb-Kangas 1984a: 90
This definition is almost the expression of an ideal. It may be at¬

tributed to the fact that (as she herself states) Skutnabb-Kangas had a
specific group in mind, namely immigrants and minority children,
‘who I hope will be given the opportunity to become so completely
bilingual that they satisfy the demands of my definition, something I
think of as entirely possible’ (my italics). Her table sets out clearly the
criteria and other details involved in the definition (see Table 1.2):
SEMILINGUALISM 27
TABLE 1.2 Defining bilingualism
Criterion The mother tongue is the A speaker is bilingual who

language
Origin first learned (the speaker has (a) has learnt two languages in the
established her first lasting family from native speakers from
linguistic contacts in) the beginning
(b) has used two languages in parallel

as means of communication from
the beginning
Competence best known (a) complete mastery of two languages
level of (b) native-like control of two languages

proficiency
(c) equal mastery of two languages
(d) can produce complete meaningful

utterances in the other language
(e) has at least some knowledge and

control of the grammatical
structure of the other language.
(f) has come into contact with another

language
Function most used uses (or can use) two languages (in most
situations) (in accordance with her own
use wishes and the demands of the
community)
Attitudes (a) identified with by self (a) identifies herself as bilingual/ with
(internal identification) two languages and/or two cultures
identity and (or parts of them)
identification
(b) identified by others as (b) is identified by others as bilingual/
a native speaker of as a native speaker of two
(external identification) languages
Source: Skutnabb-Kangas 1984a: 91
1.3 Semilingualism
A brief mention must be made of the notion of so-called ‘semilingual¬

ism’, which has been at the centre of considerable argument. Like
‘subtractive bilingualism’, it appears to suggest a negative view, to the

effect that, if a person has to use two languages, this can have det¬
rimental results on general linguistic ability, i.e. that the person is
liable to suffer from some kind of linguistic deficiency in both lan¬
guages. The expression was first used by Scandinavian linguists who
were considering the language of Finnish minority children in Sweden
(HansegSrd 1968; Skutnabb-Kangas and Toukomaa 1976, 1979). The
Swedish term halvsprdkighed was rendered into English as ‘semilin-
gualism’ or ‘double semilingualism’. It was particularly in connection
with matters related to bilingual and vernacular education that the no¬
tion became controversial. According to the observations of these
linguists, the children of Finnish immigrants in Sweden showed retar¬
dation in their linguistic ability when compared to their Swedish and
Finnish peers. They claimed that this backwardness could become a
permanent feature among these immigrants. It was argued that, if this
happened it would probably contribute to the continued stigmatization
and, as a consequence, isolation of the immigrant groups in question,
as they would be cut off from sufficiently varied linguistic input from
Finland Finns and, equally, from that of Swedish native speakers (see
also section 6.3).
The phenomenon is not a new one. When she introduced the con¬
cept in America, Christina Bratt Paulston (1974) pointed out that it had
been quite vividly described by Bloomfield in 1927 when dealing with
an Indian speaker of Menomini who had learnt English, of whom he
said that he spoke ‘no language tolerably’, and he added; ‘perhaps this
is due, in some indirect way, to the impact of the conquering lan¬
guage’. It seems that Bloomfield had some sort of purely linguistic
explanation in mind. Today one would probably prefer to try to seek
out the particular social and psychological circumstances of the case
under study, as they are often seen to affect linguistic competence.
1.4 Biculturalism
The reference to attitudes to bilingualism itself can open the discussion

to wider aspects, not strictly linguistic in nature. Language is always
used within a cultural environment, and this context tends to vary from
one speech community to another. We may take it that a native Italian
speaker is more or less conscious of Italian culture, familiar with a
whole set of shared experiences and assumptions. However, it does not
BICULTURALISM 29
follow automatically that those who know two languages are bicultu-
ral. The English-Italian bilingual from the Bronx, in New York, may
perhaps be aware of aspects of Italian culture relating to the home,
family, food, etc., but she is likely to be unaware of the popular culture
of, say, present-day Naples; on the other hand, she will be conversant
with the culture (or subculture) that is peculiar to Italian immigrants in
the Bronx. The British student of French life and letters, in contrast,
may be highly knowledgeable about French cultural affairs past and
present, yet feel insecure and be non-fluent when it comes to dynamic
oral interaction. Just as a bilingual may possess varying degrees of
competence in the two (or more) languages, (s)he may also exhibit
different degrees of biculturalism. Normally, we can expect less fluent
bilinguals to be less bicultural as well, in the same way as one would
predict that a fluent bilingual will be more familiar with both cultures;
but it will depend on the way they have acquired their languages. Mak¬
ing use again of the idea of a continuum, it is possible to imagine the
probable patterns emerging from various likely combinations of bilin¬
gualism and biculturalism. The chances are that one will find a greater
concentration of people towards the monolingual/ monocultural end of
the scale, and then a decreasing number of individuals with a high
degree of bilingualism and biculturalism, in line with the statement
previously made that ambilingualism (and equi-lingualism) are rare
occurrences. But other situations are also feasible.
Oksaar (1983) defines biculturalism (she uses the term ‘multicultu-
ralism’) on the basis of a broad view put forward by Soffieti (1955)
which sees culture as ‘the ways of a people’, i.e. the defining charac¬
teristics of a person or a group, including behaviour patterns:
‘Multiculturalism of a person is realized in his ability to act here and
now according to the requirements and rules of the cultures’ (Oksaar
1983: 20).
She states that, in the case of immigrants, the two languages usually
fulfil different roles and functions, their distribution being decided by a
number of social and psychological factors. More often than not, the
mother tongue belongs to the individual sphere and the language of the
host country to the official and sociocultural one. But, she argues, these
relations can change, and the distribution of LI and L2 in relation to
the cultural spheres may be an important criterion for the immigrant’s
degree of assimilation or isolation in the host country. One could per¬
haps add that the distribution of the two languages and cultural spheres
will not necessarily be the same for all the members of the same fam-
ily. Parents may be witnesses of how the culture of the country of

residence begins to dominate their children’s individual sphere and
how they increasingly regard the culture(s) of the parents as something
that belongs to a different sociocultural environment. This can lead to
conflicts which are, potentially, much more frustrating than they would
probably be if they were caused by language use only.
complete
biculturalism
monolingualism and
complete
monoculturalism
bilingualism
HGURE 1.1 The relationship between bilingualism and biculturalism
Many foreign language programmes of study attempt to strike a bal¬

ance between imparting ‘language’ knowledge and teaching ‘culture’,
not only because of the intrinsic value that cultural matters are as¬
sumed to have, but also because of the realization that effective
communication can proceed without hindrance only when the speakers
are aware of the cultural implications (see, for example. Byram 1986c;
1988). To apply the same set of cultural values as are used for Lan¬
guage A when addressing native users of Language B is likely to lead
A BILINGUAL PROFILE 31
to misunderstanding; the speaker may, quite unintentionally, appear to

be rude, arrogant, strange or ignorant, simply because the concordance
of linguistic items and cultural conventions is not observed. In addi¬
tion, the more fluent the bilingual becomes, the fewer allowances will
be made and the less tolerant the native speakers of the other language
will be of violations of cultural assumptions. As Baetens Beardsmore
(1982; 20) puts it; ‘the further one progresses in bilingual ability, the
more important the bicultural element becomes, since higher profi¬
ciency increases the expectancy rate of sensitivity towards the cultural
implications of language use.’
1.5 A bilingual profile
From whatever angle we look at it, bilingualism is a relative concept.

All definitions presented here go some way towards characterizing the
various forms it may take, but there is not one that is equally accept¬
able, or useful, for everybody approaching the subject. Instead of
trying to make people fit into previously established definitions, it may
be more fruitful and practical to aim at setting up a bilingual profile for
each individual (or group) which takes account of such variable aspects
as;
(1) language development (i.e. acquisition), maintenance and/or loss

of LI and L2;
(2) sequential relationship of LI and L2, i.e. whether they are ac¬
quired simultaneously of subsequently;
(3) language competence, that is, degree of proficiency in LI and
L2, and language dominance;
(4) functional aspects of language use; what, when and to whom LI
and L2 are used;
(5) linguistic features, such as code-switching, borrowing and inter¬
ference;
(6) attitudes towards LI and L2, speakers of LI and L2, and bilin¬
gualism itself;
(7) internal and external pressures (motivational, social, psychologi¬
cal, perhaps others);
(8) environmental circumstances surrounding the bilingual;
(9) biculturalism, that is, degree of familiarity with the cultures of
LI and L2.
In line with the objectives pursued in other social sciences, the ulti¬
mate goal would be to set up a comprehensive theory of bilingualism.
This is not to say that attempts at partial theoretical explanations have
not been undertaken (see, for instance, the account in Baetens Beard-
smore 1981). The more detailed bilingual profiles become, the more
insights we shall gather and the nearer we shall get to valid generaliza¬
tions on which to base such a theory.
Chapter 2
The study of bilingual children
2.1 Early and late bilingualism
The distinction between early and late bilingualism was introduced in

Chapter 1. In the following discussion the terms ‘early bilingualism’
and ‘child bilingualism’ are both used to refer to the child who has
been in contact with two languages from birth, and also to the one who
acquired a second language in early childhood, after the first language
had been established.
‘Early’ and ‘late’ bilingualism were sometimes used, in former pub¬
lications on the subject, to refer to natural or primary bilingualism and
to the artificial, secondary kind. Adler (1977) introduced the express¬
ions ‘a.scribed bilingualism’, used (mainly, but not exclusively) with
reference to children, and ‘achieved bilingualism’ to denote older indi¬
viduals. There was, however, some inconsistency in the use of all these
terms. As far as the anglophone countries are concerned, for a long
time studies of bilingualism took second place to linguists’ preoccupa¬
tion with other aspects of applied linguistics and psycholinguistics. In
fact, a fair amount of research in bilingualism was concerned with sec¬
ond-language learning and teaching in American or Canadian
school/college settings. This may help explain why the adult bilingual
was, more often than not, thought of as someone who had learnt a
second language (as different from having acquired it under natural
conditions). This assumption finds expression in the definitions of bi¬
lingualism which convey a minimalist view of language competence
(see Chapter 1).
From the end of the 1960s there was a widening of the scope of
attention to include different groups of adults and children, and mem-
34 THE STL DY OF BlUNGl AL CHILORF-N
bers of linguistic minorities (especially non-indigenous migrants aikl

immigrants such as the Hispanos in the United Stares, the 'guestwor-
kers' in the Federal Republic of Germany and other Eua''(\\in
countries, and the immigrants from the ex-colonies of the British Em¬
pire, the Netherlands. Belgium and Fnmce). Eaim that time on the
meamng of the term ‘bilingual* has been extended to include members
of immigrant or migrant communities who rarely bec:ime users of a
second language through formal teaching and le;iming metho«.is. A new
and vigorous area of research into second language acquisition spratte
up, looking into ways in which both the older and vounêr membt'rs of
these groups acquire the language of their host amntry in a ‘natural*
way, rather than learning it by formal means.* Their le;iming. or ac¬
quisition, of the second language (i.e. that of the host counm l defxnuls
on the interplay of a number of different factors, some of w hich am
examined in section 14.3 in connection with the discussion of the lin¬
guistic situation of migrant workers in West Gemtany.
So the original dichotomy ‘early*-‘late* bilingu;ilism should be
taken only as a reflection of the age of the bilingual, i.e. w hether the
invididual becomes bilingual during childhood or as ;m adult. Late bi¬
lingualism may be the result either of L2 acquisition in a natural
environment (e.g. the migrant worker from Turkey who t;tkes up a job
in Germany without any previous knowledge of Gemtan). or of second
language learning, as with the person who hits studied the L2 for years,
using graded language-teaching materials, attending courses, etc. Thus,
late bilingualism may be of the natural or the artificial kind, the pri¬
mary or the secondary, the ascribed or the achieved type. On the other
hand, early bilingualism will, in most cases, be the natural, ascribed
sort, especially in the case of the pre-school child.
There are a number of differences between early and late bilingual¬
ism; some of them will be referred to in this section. The most basic
difference lies in the nature of child language acquisition. When the
child learns to sp)eak, (s)he learns to use language as a means of ex¬
pression, communication and social contact. Tlie child acquires the
formal aspects of a language, its sounds, words, meaning relationships,
i.e. its grammar. But at the same time (s)he is also learning to use
language as a tool for understanding and manipulating the world
around her/him, i.e. (s)he is learning that (s)he needs language to form
relationships with the people who surround her or him. In other words,
language is an essential ingredient of the child’s socialization process.
First language acquisition thus differs from all subsequent language
EARLY AND LATE BILINGUALISM 35
acquisition or learning in that the young child experiences language

use, for the first time, in a social context, with the social consequences
that this fact may entail; this is the case whether one or more lan¬
guages are being acquired from birth. The learning processes involved
in late bilingualism can draw on the social and communicative experi¬
ences gained in childhood. This represents a considerable, yet often
underestimated, advantage enjoyed by the older learner. Language pat¬
terns and assumptions about linguistic usage which have been acquired
in the mother tongue are likely to help the learner when coming into
contact with a new code, as he/she will extend them by analogy -
although the other side of the coin is that this habit may result in inter¬
ference when the two systems diverge.
2.1.1 Is early bilingualism ‘better’ than late bilingualism?

There is a widespread belief that equates child bilingualism with ‘true’
bilingualism, and some specialists seem to adhere to this view. For
example, Adler (1977: 13) says: ‘One fact is clear: whether a person in
his future life really masters two languages completely is decided in
early childhood. When he does not learn the languages then he will
never be completely perfect in both.’ Two points need to be addressed
here: one, the idea that the person ‘will never be completely perfect in
both’; the other is the assumption that children per se have better lan¬
guage learning abilities than adults.
With regard to the first point, it is unrealistic to suggest that the
bilingual speaker achieves complete, 100 per cent mastery of two lan¬
guages (ambilingualism). The total linguistic repertoire of a fluent
bilingual consists of items from both languages which complement
each other and may overlap to varying degrees (see Chapter 4). As far
as the second point is concerned, it is not possible to find solid proof
that children are better than adults at acquiring a second language. It is
possible to get the impression that small children achieve fluency in a
second language more quickly than older people, but this idea is not
supported by hard evidence, simply because one is not comparing like
with like. The apparent ease with which a child acquires a second lan¬
guage and the greater analytical abilities of the older learner are
important factors which are simply not comparable.
Although we can establish that there is a fundamental difference
between the processes involved in first language acquisition and sub¬
sequent language learning, this does not mean that the difference is of
36 THE STUDY OF BILINGUAL CHILDREN
either a qualitative or a quantitative kind. The only exception may be

pronunciation. Children are believed to have greater phonetic-auditory
ability, which enables them to distinguish and reproduce new sounds
quite easily, whereas adults may experience some degree of inter¬
ference from their LI (more about this later in the chapter).
2.1.2 The Language Acquisition Device (LAD) and the Critical

Period Hypothesis
Two ideas taken from first language acquisition are relevant in this
context. Both have had considerable influence on linguistic thinking
and on the belief that children have better language learning abilities
than adults. In 1959, in a famous review of the book Verbal Behaviour,
by the psychologist B. F. Skinner, Chomsky expressed his conviction
that children acquire language with amazing speed and efficiency, and
that this cannot be accounted for ultimately in terms of stimulus and
response, as suggested by Skinner, an exponent of the Behaviourist (or
‘mechanistic’) theory. Chomsky cites the example of the immigrant
child who has no problems with the acquisition of his new country’s
language, whereas his parents struggle with it, learn less successfully
and show strong signs of phonological and syntactic interference from
their native tongue in their use of the second language. In the 1960s it
became a main pillar of ‘mentalist’ theories of language acquisition to
postulate that children are somehow specially pre-programmed, or pre¬
disposed from birth, to acquire language. The ‘Language Acquisition
Device’ (LAD) was the expression used to refer to a hypothesized in¬
nate gearing towards language, which needed only to be activated to
start functioning, i.e. the child just had to come into natural contact
with a human language for the LAD to be set off.
Some research into psychological and neurological aspects of lan¬
guage appeared to back up the idea that there is a ‘critical period’
during which children are particularly adept at acquiring language
(Penfield and Roberts 1959; Lenneberg 1967), which came to be
known as the Critical Period Hypothesis. This period was supposed to
last from about the second year to the age of puberty. There was said
to be a biological link to the development of the brain’s dominance of
language function through lateralization, i.e. the specializing of one
side of the brain (usually the left) in dealing with language. Before the
age of two language acquisition is not possible, it was thought, because
of maturational factors, and after puberty the brain loses its plasticity.
EARLY AND LATE BILINGUALISM 37
i.e. is no longer particularly receptive for the task.

The suggestion of the critical period has been reviewed by a number
of psycholinguists (see McLaughlin 1984 for a thorough account), and
both the upper and lower age limits have been questioned. In the early
1970s it was suggested that lateralization might be complete long be¬
fore adolescence (Harshman and Krashen 1972; Berlin et al. 1972),
and possibly essentially complete at birth (Krashen 1975). Seliger
(1978) proposed the idea of different critical periods for distinct abi¬
lities as a way of explaining why a number of aspects of language can
be acquired at varying ages. The relationship between a presumed criti¬
cal period and lateralization has not been satisfactority dealt with so
far. The notion therefore remains very much a hypothesis (see also
section 11.2).
It is possible, however, that there is a period during which a child
has a special facility for neuro-muscular patterns, i.e. during which
(s)he finds it particularly easy to acquire any pronunciation features.
There is some evidence (many of the authors of the case studies men¬
tioned in section 2.4 later, for example, make this point) that supports
the popular impression that children are good at acquiring a native-like
accent in a second or foreign language, but the issue has by no means
been established beyond doubt. What is generally accepted, though, is
that children have certain qualities that favour second language ac¬
quisition: they make good mimics, they lack some of the inhibitions
that get in the way of many adult language learners, and they have a
greater capacity for learning by playing. All this can positively affect
their fluency and pronunciation.
2.1.3 Children are not necessarily better bilinguals

Chomsky, as we have seen, was impressed by the apparent speed and
efficiency with which children acquire language. But closer investiga¬
tion of the issue, it has been noticed more recently, can produce a
different impression, even the opposite one. If we suppose that the ac¬
quisition process starts at birth, as the infant is exposed to language
from that moment, we can claim that children actually spend a long
time hearing language, before they start using it actively at around the
age of two. Furthermore, to say that language development is complete
by the age of five or six is something of an exaggeration. On the other
hand, the linguistic standards expected from a child are generally much
lower and less sophisticated than what it is hoped any adult will pro-
duce. In fact, it can be maintained quite reasonably that it is the adult

who appears to learn fast and to master quite a wide range of language
uses in relatively little time, when one considers the time taken in first
language acquisition. Observation of young language learners can also
indicate that children are rather unsophisticated in their learning pro¬
cess, as they lack a number of skills which the older learner usually
has, and which can facilitate (for the latter) transfer from one language
to another: children do not normally perceive the similarities between
languages, and they are unable to abstract, classify and generalize to
the extent that adults can.
For some time there was support from certain academic fields for
the idea that children’s language aptitude is, in general, superior to that
of older people. Perhaps as a consequence of this, and in analogy to
such belief, there was a widely held view that children make better
bilinguals than adults. But as no scientifically based evidence sugges¬
ting that there is a biological basis for a critical period has been put
forward, we cannot say that children have an intrinsic language ability
of a superior order, with the possible exception of a phonetic-auditory
ability. The child may well be linguistically more adept with respect to
the acquisition of the phonological system; and certain psychological
factors (relating to favourable disposition towards mimicry, playfulness
and lack of inhibiting barriers) may facilitate early fluency. But that is
all tlât we are entitled to claim.
Apart from this possible advantage, then, children cannot be said to
be better bilingual learners. What is more, adults possess a number of
analytical skills that can stand them in good stead when learning a
second language. In view of all this, the successful establishment of
bilingualism may well depend on psychological factors (such as atti¬
tudes, motivation and willingness to identify with the speakers of the
L2), rather than physiological or biological ones. If this is so, it will
apply to children as well as adults.
2.2 How does the child become bilingual?
2.2.1 Some examples

The obvious answer to the question of how the child becomes bilingual
is: by growing up in a bilingual environment. But what constitutes a
bilingual environment? Why do some children appear to become bi-
HOW DOES THE CHILD BECOME BILINGUAL? 39
lingual almost spontaneously, while others seem to need extra help and
encouragement? Grosjean (1982: 167ff) mentions the example of two
children of the same age, who live in the same American city, both
bilingual, but who have become so by very different routes. Ingrid, the
daughter of an English-speaking American father and an Swedish
mother, was brought up to speak the two languages from birth; both
parents were professional people, they spoke their respective native
languages to their daughter consistently and made conscious efforts to
provide her with rich and varied linguistic stimuli. In the case of Swed¬
ish (i.e. the language which was not the one in everyday use around
Ingrid outside the home) this meant books, cassettes and other material
in Swedish and as much input as possible from Swedish-speaking
friends and relatives. When Ingrid entered school she was a happy and
bright child who was a fully functional bilingual.
The case of Dieudonne was quite different. When he was five he
moved with his parents to the United States. He spoke Haitian Creole,
and after their arrival his family continued to speak this language to
him. Since both parents went out to work, Dieudonne spent a good
deal of time with his brothers and sisters, or playing with other child¬
ren, many of them Puerto Rican, Black American and Haitian, and he
picked up some English from them (and other English-speaking
friends) and from television programmes, but not enough to be placed
in a mainstream class when he started to attend school. He joined,
together with other minority children, a programme in which English
was taught as a second language, and he experienced many problems
there, as he was unhappy and withdrawn and was said to be making
little progress. After a frustrating period he was transferred to a school
where he was taught in Creole, and where the other children also used
this language. English was introduced slowly, and then progressively
used as the medium of instruction. Eventually he went on to main¬
stream American education, his bilingualism having been firmly
established.
Graciela is a recent acquaintance of mine. She is Spanish, lives in
Madrid and is about to start at university. Her mother speaks Spanish
only. Her father was bom and grew up in Catalonia, speaking both
Catalan and Spanish. Although he does not use Catalan with his child¬
ren, the family spend most of their holidays in a Catalan-speaking
environment in which some of the relatives insist on addressing
everyone in Catalan only; so Graciela has, over the years, gained
enough undertanding of spoken Catalan to be able to converse fluently.
understand TV programmes and generally cover her needs in this lan¬

guage, although she does not write it and prefers reading books,
newspapers and magazines written in Spanish. For her entire schooling
Graciela has attended an English school in Madrid. Almost all the staff
are native speakers of English, and for most of her time there, particu¬
larly in the lower school, English was used as the medium of
instruction, although the children would use Spanish among themselves
and either English or (some of the time, when in the upper school)
Spanish with the teachers. Graciela has no problem understanding
spoken or written English, she can write it quite well and is able to
speak it with confidence, although she does not sound like a native
speaker as she lacks this kind of fluency and her pronunciation is not
faultless.
There are many children in the world who, like Ingrid, Dieudonne
and Graciela, have become bilingual, or even trilingual, at some earlier
or later stage of their childhood, some under more favourable circum¬
stances than others. Clearly, there can be a multitude of different ways
in which children may achieve bilingualism. A number of them have
been studied in some detail, as we shall see later, others have been
described in books written by parents or linguists (sometimes both
coinciding in the same person) and published as ‘guides’ or ‘hand¬
books’ for interested lay persons, parents, teachers and others involved
in the care of bilingual children (e.g. Saunders 1982; Kielhbfer and
Jonekeit 1983; de Jong 1986; Harding and Riley 1986; Arnberg
1987.^
2.2.2 Bilingual patterns

The ways in which children can become bilingual, and their bilingual¬
ism be maintained, vary a great deal from one family or individual
case to another. In this section some commonly found patterns are out¬
lined.
(1) Immigration Immigration involves leaving the country of origin in

order to settle, once and for all, in a ‘host’ country. Nations like the
United States and Canada have, over the centuries, seen large numbers
of immigrants entering their territories. Britain, France, Belgium and
the Netherlands have also seen considerable quantities of immigrants
settling in their lands, after the emancipation of their colonies. The
children of immigrants usually acquired their first language at home.
from their parents and family, and their second, that of their new
country of residence, from people outside the home. In the majority of
cases there was no special provision made for them when they entered
school, so they had to ‘sink or swim’ as best they could. In most coun¬
tries immigrants were expected by public opinion to ‘assimilate’ (i.e.
adapt to the norms and customs of the host country) as quickly as
possible, and this of course included linguistic assimilation. The result
usually was that, for many generations of immigrants, bilingualism was
a transitory stage, lasting only a limited number of years:
First Second Third Fourth

generation generation generation generation
LA-i- LB LA and LB +LA, LB LB
(+ = ‘some’; LA = Language A; LB = Language B)
(2) Migration Many western European countries experienced large-

scale migration in the three decades after the Second World War, as
people moved across frontiers in search of work and better living con¬
ditions. In some cases they took their families with them from the
outset or sent for them later, in others they found partners from the
host country or from other ethnic groups. The children of migrants
may grow up hearing the language of their parents only, if they live in
a community composed mainly of migrants of homogeneous origin,
and their contact with the second language may not begin in earnest
until they are of school age. Their experiences may be similar to those
of the child mentioned above, Dieudonne. On the other hand, they may
mix to a larger extent with the children of the host country, go to a
kindergarten or playgroup where the local language is used - or they
may learn it from one of their parents. In other words, whether second-
generation migrant children become bilingual or not depends very
much on family circumstances and the surrounding social conditions
that affect them. Migration is seen as involving temporary movement
only, i.e. an eventual return to the country of origin is contemplated.
For migrants it is therefore important to maintain their language, but
by the same token the host country may feel relieved of any obligation
to make special educational provision for their children, with respect to
both distinct second language programmes and mother-tongue teach¬
ing.'^ The ways in which these children become bilingual, if at all, are
more unpredictable than in the case of immigrants.

Lewis (1972), in his interesting account of multilingualism in the
Soviet Union, includes a discussion of the development of what the
author calls ‘mass’ or ‘popular’ bilingualism. Movement of people
from one Republic to another and consequent inter-ethnic marriage are
apparently quite common in the USSR, and many children become bi¬
lingual as a result. Another way of becoming bilingual in the USSR is
through the school system, as educational planning includes a provi¬
sion for the teaching of, and in, Russian at school in
non-Russian-speaking areas. Many Soviet children may, therefore, ac¬
quire either one or two languages at home (from their parents), and
another language during their schooling.^
(3) Close contact with other linguistic groups In some multinational

states or countries with rich linguistic diversity, contact between mem¬
bers of different language groups is quite common. It may be brought
about by urbanization or by internal migration, and bilingualism is
likely to be found among children as well as adults. The children may
have parents or other family members who speak different languages,
or they may hear one language at home and another one outside it. In
Europe this kind of scenario is usually found among minorities (e.g. in
Wales, the Basque Country, South Tyrol or Friesland). This element of
dominant majority and subordinate minority, however, is absent in
Mkilifi’s (1978) analysis of the language situation in Tanzania. Many
of the children he studied had, before entering school, acquired Swahi¬
li, in addition to the local language, in part from the father (usually
from him, as the men tend to have more outside contacts through their
work, and therefore usually know Swahili better than the women) and
partly from other members of their community who would frequently
use it as the means for wider communication. Mkilifi reports that bil¬
ingualism is a widespread and natural outcome in these circumstances,
and that the children he observed were often not even aware that they
spoke two different languages.
(4) Schooling Nowadays education can play a very important role in

making children bilingual. The educational system may deliberately be
geared towards fostering bilingualism, as is the case, for instance, with
the French immersion courses in Canadian schools, designed for child¬
ren from the English-speaking majority, and also, increasingly, with
Catalan courses for Spanish-speaking children in Catalonia. Another
example is the compulsory use of a second language as a medium of

instruction at the level of secondary education (or earlier, in some
parts) in the USSR, where, as Lewis (1972) points out, three important
linguistic changes have been brought about since the introduction of
universal education: first, the development of mass bilingualism,
whereas previously there was only a limited amount; second, the fact
that this so-called ‘popular’ bilingualism is now of a literate nature,
while formerly it was mainly oral; and third, that becoming bilingual is
now an option open to the vast majority of Soviet children and no
longer the prerogative of an elite.
The various International Schools, the European School (as dis¬
cussed, for example, in Baetens Beardsmore 1979; 1980), and the
American, English, Erench, German and Spanish schools in a number
of European and other countries, also contribute towards establishing
bilingualism, as does the practice of having one’s children educated in
a different country, e.g. German or Spanish parents sending their child¬
ren to a British boarding school, in the western European context. The
numbers involved under this last heading are not very significant, as
these schools are accessible only to the children of highly mobile and
well-off families. It is clear, therefore, that in such cases we are, in
fact, talking about elite bilingualism (see section 2.3).
Only a few European minorities have obtained the right and the
necessary financial provision to have their children educated through
the medium of the minority language, even though it is quite clear that
without such support a large number of minority languages are likely
soon to be lost. The Welsh had to wait until the late 1960s to see the
establishment of the first Welsh-medium schools. Some would argue
that too little has happened too late for the decline of the Welsh lan¬
guage to be halted. Schooling in the non-mainstream language can
assume paramount importance for the survival of a linguistic monority.
In a perceptive study of the German minority in Denmark, Michael
Byram (1986b) shows that the role of the German school goes far be¬
yond the provision of a German education for its pupils and a focal
point for the community: ‘Although at first sight the minority is ‘Ger¬
man’ because of the language and culture it possesses, more careful
observation suggests that the existence of the minority depends on its
schools.’ (Byram 1986b: xii).
More often than not, at any rate in western Europe, the education
system reflects the one-state-one-language principle followed for a
long time by many countries. Children from linguistic minorities are
treated in the same way as those belonging to the majority group,

which means that no provision is made with regard to their special
linguistic needs. In the past many schools in Europe have pursued
strict anti-minority policies, even to the extent of punishing or ridicu¬
ling children who used the minority language at school. The repressive
methods employed in places such as Gaelic-speaking Scotland, Wales,
Brittany, Catalonia and Galicia (in north-west Spain) were remarkably
similar and, sadly, usually Just as effective. Because there was no sup¬
port for the minority language, children had to acquire the language of
the majority as well as they could under difficult circumstances. What
bilingualism existed was often of a transitory nature, and the shift to¬
wards the majority language took only one or two generations. Today
we see less overt oppression of linguistic minorities. In the most fa¬
vourable cases the schools offer some kind of programmes in the
child’s mother tongue (as happened with Dieudonne), or there are spe¬
cial classes to help children learn the second language before they join
the main body of the education system, e.g. ESL (English as a Second
Language) provision in the UK or ‘Deutsch als Zweitsprache’ classes
for migrants’ children offered in some states of the German Federal
Republic. There are also certain private schemes which aim at teaching
minority languages to children already in mainstream education, for
instance in the Scandinavian countries, Gennany, France and Britain.
But, as participation is voluntary, their success must depend on such
factors as motivation, resources and length of attendance.
(5) Growing up in a bilingual family At the family level there are

many different strategies to choose from for bringing children up bi-
lingually. Many of the children observed in the case studies listed
below (see 2.4) came from families where one parent spoke the lan¬
guage of the wider community and the other parent a ‘foreign’
language. This was the case of Danny, a Swedish-English bilingual,
the father being Swedish, the mother American, living in Sweden; or
of Lisa and Giulia, Italian-German bilinguals, whose father is Italian
and mother German-speaking, who live in Italy. In such families the
one-parent-one-language principle was adopted and bilingualism estab¬
lished. The degree of success will depend mainly on such factors as
whether the parents are consistent in their language use, whether the
child has enough exposure to the ‘home-only’ language, whether (s)he
perceives the need to use both languages, and whether (s)he receives
the right kind and amount of social support. Providing the first two
conditions, consistency and exposure, are met, the establishment of bi¬

lingualism is not usually problematic. However, the maintenance of the
‘home-only’ language, at an early age as well as later, is much less
certain. Some children may not see much point in using the language
of one parent, once they realize that he or she also speaks the language
of the other. This happens quite often, particularly when the latter is
also the language of the society where the family lives. When the child
is older the input from the weaker language may be too one-sided, too
limited in register and style (when compared with the rich stimulus in
the outside language, which is of course also used at school) for the
two languages to develop on the same level. In fact, judging from
parents’ accounts (and my own experience), bringing children up bi-
lingually requires considerable effort, expense and ingenuity. There are
many families who cannot afford such luxuries as frequent visits
abroad or the purchase of books, cassettes, videos and films, quite
apart from the time needed to use these with the children.
In those families where a minority language is spoken by one or
both of the parents, the aim may be either to introduce the two lan¬
guages to the child from birth, or to begin with the majority language
after the minority one has been established at home but before the
child starts school (where the latter will be used). The successful devel¬
opment of bilingualism will depend on the same considerations as
before: exposure, consistency, perceived need and social support from
both majority and minority communities. But social support is likely to
assume increased importance. Whether the attitudes of the majority to¬
wards the minority (and vice versa) are positive or negative, and
whether the knowledge of the minority language is considered an asset
or not will be very influencial factors.
Not all families opt for a consistent pattern of language use; nor do
they always adhere to the one-parent-one-language principle. The par¬
ents, and other family members, may use both languages; they may
apply, perhaps inconsistently, practices such as one parent using one
language when alone with the child and the other when other family
members are present; or they may follow no specific pattern at any
time. For the establishment of bilingualism this kind of strategy tends
to be less successful, as then the choice of using a particular language
at any given time will depend on arbitrary factors, and the child may
find this confusing. If this happens, the majority language may soon
become the dominant one, and the incidence of mixed language output
is likely to be high.
There are many possible answers to the question of how children

become bilingual; only a small proportion of them have been men¬
tioned here. Individual family circumstances, as well as those
prevailing in the wider community, may decide whether, and to what
extent, a child becomes and stays bilingual, and a number of psycho¬
logical, social and educational determinants will come into play also
(see further discussion of this in Chapters 6 and 7).
2.3 Types of bilinguals
Skutnabb-Kangas (1984a) suggests a classification of the world’s bi¬

linguals into four groups. In drawing distinctions between these four
categories she takes the following factors into account:
(a) pressure to become bilingual;

(b) the prerequisites for bilingualism;
(c) route by which the individual has become bilingual;
(d) the consequences entailed in failing to become bilingual.
She identifies the groups as follows:
(1) Elite bilinguals These are people who have freely chosen to
become so (e.g. because they want to work or study abroad), and child¬
ren who belong to families who change their country of residence
relatively often and/or who are sent to be educated abroad. In these
cases, the normal situation is that the acquisition of both languages
proceeds unhindered, with the two languages receiving wide social
support and the mother tongue, in particular, enjoying a firm and stable
position. The second language may have been either learnt or acquired
but, as the attempt to establish it firmly will have been quite voluntary,
failure to gain sufficient command of it (if failure there is) will carry
no serious consequences for the subject. Children who live temporarily
in a different linguistic environment may feel a greater need to learn
the language of the host country in order to make social contacts or be
able to follow the school curriculum. Their attempt to make progress in
the L2 will usually be met at the very least with friendly approval, and
they will confidently expect that one day they will return to the
country of their mother tongue.
(2) Children from linguistic majorities These are children who learn
another language (e.g. that of a minority group) at school, such as in
TYPES OF BILINGUALS 47
immersion programmes or in foreign language classes. The learning of

the second language may, for example, be considered advantageous
either because it is seen as a way of enhancing the prestige of the
minority language (as with French in Canada) or because it is believed
to be of wider educational or vocational benefit (for instance, English
as a foreign language in the Netherlands or in many other countries).
Usually majority children experience little or no pressure (from family
or society) to become bilingual; a variety of well-prepared materials,
designed especially for them, will facilitate the learning; and the risk
involved in failing to achieve the learning objectives tends to be rela¬
tively small.
The children from these two groups tend to come from monolingual
backgrounds, in contrast with the following two.
(3) Children from bilingual families These are children whose parents
have different mother tongues. The child will experience considerable
societal pressure to become fluent in the official language, but there
will be no external compulsion to become bilingual. Bilingualism will
be desirable because (and to the extent that) there are internal family
pressures requiring the child to communicate in the language of the
parent(s). The emotional relationship between the child(ren) and the
parent(s) may suffer somewhat if bilingualism does not develop. But
for the child’s educational success and complete social integration the
only important thing, ultimately, is that (s)he acquires the language of
the country. So the consequences of failure to become bilingual may
possibly be problematic within the family, but not too serious at the
societal level.
(4) Children from linguistic minorities These children have parents

who belong to a linguistic minority; they are under intense external
pressure to leam the language of the majority, particularly if the lan¬
guage of the minority is not officially recognized. Yet there may be
little support offered to them in terms of bilingual education pro¬
grammes or primary/secondary school teaching. They often find
themselves also under the influence of strong internal forces that en¬
courage them to leam the language of their parents and to form social
relationships with the members of the wider (minority) group. The risk
of failing in the attempt to become bilingual is greater than for any of
the above groups. As Skutnabb-Kangas says, the effects of lack of suc¬
cess in attaining bilingualism may be catastrophic, ranging from loss of
educational and future opportunities to problems of rootlessness and

alienation.
2.4 Case studies of bilingual children
Much of what is known today about bilingual language development is

based on the findings presented in case studies; some of them are listed
below. Of these, a few are quite well known, above all perhaps the
work of the German-bom linguist Leopold, who lived in the United
States and studied in detail the English—German bilingual development
of his daughters Hildegard and (later) Karla over a period of some ten
years. His diary entries and analyses have become a standard work of
reference for any student of child language development, not just bi¬
lingual development. A number of these case studies are reviewed by
McLaughlin (1984: 88) and Skutnabb-Kangas (1984a: 146-9).
In many cases the children studied are the linguists’ own. In terms
of the prevailing socio-economic background, one imagines that these
children could be classed as elite bilinguals, as they all received a good
deal of support from inside, and more often than not also from outside,
the family. Thus becoming bilingual was not absolutely essential for
them. Tabouret-Keller is an exception: he studied a French/Alsatian¬
speaking child with working-class parents. Some of the purely
linguistic observations made in these case studies can be taken to be
applicable to children from a wider range of backgrounds. But gener¬
alizations about the psychological, social and educational factors
concerning the establishment of bilingualism should be made with a
great deal of caution.
Quite often these studies make fascinating reading, as they provide
a wealth of linguistic and other detail, and also many insights about the
cognitive development and family life of the bilingual child. They are
usually based on carefully planned and well-conducted surveys or col¬
lections of diary entries and recorded data, sometimes supplemented by
anecdotal evidence. There is even the case of the linguist who reports
on his own bilingualism (Elwert 1973). However, all such studies
should be approached as critically as possible, because they may suffer
from certain shortcomings.
Before cassette and video recorders were available, data had to be
transcribed or taken down in longhand. Problems of reliability have not
been totally solved by technology. We can now record children’s
CASE STUDIES OF BILINGUAL CHILDREN 49
speech, but unless the material is transcribed (itself an extremely time-

consuming exercise) very soon afterwards, it may be quite
unintelligible for the researcher at some later stage. Adults may be very
good at understanding what children mean when they experience it in a
communicative situation, but if it is taken out of its context vital clues
can easily be lost. A less obvious drawback is that data might be re¬
stricted to one or two children, often the eldest child in the family, so
the findings may have less bearing on bilingual development among
the younger siblings or on the question of differences across several
children within the same group. A further issue, in particular with re¬
spect to some of the earlier case studies, is that they tried to cover a
large number of developmental aspects in a relatively short period. The
more recent studies include a greater amount of detail, and they pro¬
vide information not only on the acquisition of the linguistic systems
but also on such aspects as general communication ability, biliteracy
and attitudes (e.g. Saunders 1982a; Taeschner 1983; Fantini 1985). We
are just beginning to see the emergence of longer-term studies, extend¬
ing over the whole of infancy and childhood (for instance, Saunders
1988, which was a study of developing bilingualism up to adoles¬
cence). This is an important step forward, as little is yet known about
such issues as how older children maintain their bilingualism, what
effect school has on it, whether biliteracy is attempted, and with what
result, etc.
Another welcome development is the initiation of research projects
involving the bilingual development of several children (e.g. Redlinger
and Park 1980; Oksaar 1980; Meisel 1984; Kutsch and Desgranges
1985).^ This kind of research allows comparisons between children,
and also across languages, under more controlled conditions, which
makes it likely that the results will be more reliable. It is, of course,
possible to draw comparisons between the various case studies. Some
attempt at this is included in the discussion of patterns of bilingual
language acquisition (Chapter 3). But one must not lose sight of the
(perhaps severe) limitations on validity imposed by the facts that a
large number of uncontrollable variables are involved in individual
longitudinal case studies and that many of these variables (e.g. those
relating to language input) have not been acknowledged or considered.
In addition, no two studies have applied the same types of tests to
measure the degree of bilingualism or of relative competence in each
language involved. Perhaps these disadvantages can be avoided to
some extent in the newer kind of studies. Another advantage of the
latter is that they allow the team of researchers to look at specific as¬
pects of bilingual acquisition, such as language differentiation, code¬
switching or the acquisition of particular syntactic elements such as
negation or word order. Perhaps these studies, and others, possibly of a
less longitudinal nature, involving groups of children, will eventually
throw some light on questions about which at the moment we know
little for certain: e.g. why some children are more successful than
others in the same family in becoming bilingual, or whether there are
cmcial moments or stages in their linguistic and cognitive development
at which certain things should, or should not, happen.
TABLE 2.1 Some long-term studies of simultaneous language acquisition in

bilingual children
Author Languages Age at time Concerns Number

and date of study of study of children
Ronjat French/German first 5 yrs P, Se, I 1

(1913)
Pavlovitch French/Serbian first 2 yrs P, Se, F, 1

(19^0) I, Mix
Hoyer Russian/German first 12 B 1

and Hoyer months
(1924)
Smith Fnglish/Chinese first 4 yrs Se, I, Mix 8

(1935)
Emrich German/Bulgarian first 3 yrs I 1

(1938)
Leopold* Fnglish/German first 2 yrs B, F, Se, 2

(1939, (and 2 to 12) DS, Si, M,
1947, OP, F, I
1949a, 1949b)
Burling Fnglish/Garo 1:4 to 2:10 P, Se, S, 1

(1959) M, I, Mix

Tabouret-Keller French/Alsatian 1:8 to 2:11 Se, I 1

(1962)
Raffler-Engel English/ Italian first 4 yrs P,Se, 1

(1965) M, I
Murrell S wedi sh/Finnish/ 2:0 to 2:8 F, Se, DS, S 1

(1966) English M, I, WO, I, Mix
Imedadze Russian/Georgian 0:11 to 3:0 Se, E, S, M 1

(1960, DS, I
1967)
Rulê-Draviii,a Swedish/Latvian first 6 yrs P, Se, 2

(1967) DS, I, Mix
Mikes Hungarian/Serbo-Croat first 6 yrs S, M, DS, 3

(1967) I, Mix
Oksaar Swedish/Estonian 0:2 to 3:0 F, Se, E, S 1

(1970) M, I, Mix
Swain$ French/English 2:4 to 4:0 S, M, DS, I 1

(1972)
Volterra and Italian/German 1:0 to 3:0 Se, E, S, 2

Taeschner M, I
(1978)
Itoh and Hatch+ Japanese/English 2:6 to 3:1 Pe, Se, S, 1

(1978) M, 1, Mix
Celce-Murcia French/English 2:4 P, Se, S, 1
(1978) M, I, Mix
Redlinger German/Spanish 2-3 yrs Se, M, 4
& Park"^ German/English (for 5-8 I, Mix

(1980) German/French months)
Amberg Swedish/English first 5 yrs M. LL 4
(1981)

Cunze German/Italian first 5 and M, S, DS 1

(1980) a half yrs
Vihman English/Estonian 1:1 to 2:10 P, Se, S, 1

(1982) M, I. Mix
Saunders* English/German first 13 yrs M, S, 2

(1982a, 1988) WO, I
Taeschner* German/Italian first 8 yrs P, Se, 2

(1983) S, M, WO,
DS, I
Fantini* English/Spanish first 11 yrs P, S, M, 1

(1985) WO, Mix
Vila Catalan/Spanish 1:2 to 3:2 Se, E, DS 3

(1984)
Hoffmann English/German/ first 8 yrs P, S, M, 2

(1985) Spanish DS, I, Mix
Meisel + German/Erench 1:0 to 4:0 M, S, 2
(1984'^'1987) WO, I
De Houwer* English/Dutch 2:7 to 3:4 M, S 1

(1990)
Stavans§ English/Spanish/ 5.5 to 6.11 c 2

(1990) Hebrew and 2.6 to 4.0
Notes
$
unpublished PhD thesis (entitled ‘Bilingualism as a First Language’)
these studies are cross-sectional rather than longitudinal
* studies that have been published in book form
§ unpublished PhD thesis
Key
B = babbling P = phonology M = morphology

S = syntax Se = semantics I = interference
DS = developmental sequence WO = word order
LL = language loss E = extension of meaning C = code-switching
Ayse Oktem and Jochen Rehbein, of the Centre for Research into
Multilingualism at Hamburg University, West Germany, have com¬
piled a useful annotated bibliography of child bilingualism. They list
over seventy longitudinal and cross-sectional studies of bilingual child¬
ren and include a bibliography consisting of some 550 entries. It was
published in 1987.
Notes
1. Incidentally, the term often used by German scholars, ‘ungesteuerter

Zweitsprachenerwerb’, is more explicit than ‘second language ac¬
quisition’, as it includes reference to the unstructured, untutored
nature of the language acquisition process in these cases. The Eng¬
lish convention is to use ‘acquisition’ in the context of first
language development and ‘learning’ for the addition of a second,
or third, etc., language in adolescence or adulthood as a result of a
conscious learning effort.
2. On the other hand, language teachers can quote many examples of

adults who have had complete success in mastering a new language
fully, inclusive of its pronunciation system, just as there are many
immigrants about whom the same can be said. It might be relevant
to ask by whose standards the pronunciation is considered to be
‘perfect’ - the phoneticians’, the language teachers’ (or some of
them), the native speakers’? And what criteria would one apply in
Judging a ‘native’ accent - impressionistic comparability with the
(very well) educated person’s accent, the possibility of taking the
person for an ordinary native speaker purely on the basis of the
accent (in these cases, who would be a reliable judge?), or some
other basis?
3. From 1983 onwards a Bilingual Family Newsletter, for bilingual

families throughout the world, has been published by Multilingual
Matters, a British publishing firm, edited by the Australian, George

Saunders, with the aim of supporting bilingualism within the family
and enlightening the non-specialist on matters bilingual.
4. As a result of EC legislation, as well as individual and national
initiatives, the situation in western Europe is undergoing certain
changes. Many countries now offer mother-tongue programmes
and/or special instruction in the school language, so as to enable
non-native children to be integrated into state mainstream education.
However, there are vast discrepancies in the resources allocated for
these children’s education among EC members, and hardly any
commonly agreed policies so far.
5. Enforced mass bilingualism has great administrative advantages for

a multinational state such as the USSR. However, it is also a poten¬
tial liability. When the people involved see their national identity
and language threatened as a result of one-sided policies they may
withdraw their support for the system, or even rebel against it, as
events in the late 1980s in the Baltic states, Armenia, the Ukraine
and Belorussia, have shown.
6. The studies by Redlinger and Park concern children who were bi¬
lingual in different languages, although they all shared German.
Those by Meisel relate to Erench-German bilinguals (this is an on¬
going project which so far has reported findings on language
differentiation and the acquisition of case markers and word order).
The children described by Oksaar were bi-, tri- and quadrilingual in
German, Swedish, English and Estonian, and they lived in Ger¬
many, Sweden and Australia; twenty-one children from different
social backgrounds were involved. Kutsch and Desgranges report on
a longitudinal study of some thirty Turkish children acquiring Ger¬
man.
Chapter 3
Patterns of bilingual language acquisition
3.1 The description of bilingual language acquisition
This chapter aims at presenting a general overview of the ways in

which children can acquire two (or more) languages. The data men¬
tioned here are taken from some of the studies listed in Chapter 2. In
virtually all cases the children studied acquired their languages simul¬
taneously from birth. But those children who acquired more than two
languages (e.g. Oksaar 1970; Murrell 1966; Hoffmann 1985) were not
necessarily infant bilinguals in all their languages. For instance, my
daughter Cristina had already started to speak Spanish and German
when she came into contact with English at the age of two and a half.
Her brother, three years younger, did hear English spoken around him
(e.g. from TV, radio, his sister’s friends) from birth, although English
was not spoken in the home by any of the family members. So Cristina
could be classed ‘infant bilingual’ in Spanish and German and ‘child
bilingual’ with respect to English, whereas in Pascual’s case the dis¬
tinction is less clear-cut; one might call him an ‘infant trilingual’,
although the development of his three languages did not proceed in a
parallel fashion. It is quite clear that the individual circumstances of
bilingual language acquisition must be at least slightly different in the
case of almost every infant/child bilingual/trilingual, and that many of
them will not fit perfectly into any descriptive category.
The description of bilingual acquisition outlined here follows the
sequence of child language development only loosely. In the one-word
(or holophrastic) stage one can already observe features of the devel¬
oping phonological and lexico-semantic systems. As the child’s
utterances increase in length and complexity, it becomes possible to
56 PATTERNS OF BILINGUAL LANGUAGE ACQUISITION
observe the emerging syntax and the ways in which (s)he dilferentiates
the two codes and keeps them separate. Phonological and lexical de¬
velopment will be discussed here before dealing with the acquisition of
morphology and grammar. However, this should not give the impress¬
ion that language development necessarily follows a linear progression.
Language operates on two levels, so that meaningless units of sound
combine together to form units of grammar (morphology and syntax)
which have meaning. Learning to speak thus involves the simultaneous
activation, and evolution, of a number of different processes.
3.2 Phonological development
From the more recent research into first language acquisition we know
that a child’s receptive language skills begin to develop in very early
infancy - hence the recommendation that bilingual upbringing should
start from birth. The newborn baby reacts differently to human and
non-human sounds, and it soon begins to distinguish pitch and stress
features. The recognition of features such as vowel length and quality,
or friction, plosion and nasality, require an analytical ability which
operates regardless of the specific language input. Similarly, the pro¬
cess involved in producing the first speech sounds in the child’s
holophrastic stage largely follows the same route in bilinguals as in
monolinguals. Fantini (1985) reports numerous entries in the diary he
kept on his son Mario’s Spanish development, which illustrate early
recognition and differentiation of sounds. At the age of 0:4 (four
months) the boy is said to have been able to recognize his parents’
voices, and by age one he was producing sounds which followed fam¬
iliar intonational patterns, even if the sounds themselves were
meaningless. When he was 1:10, at a party surrounded by people all
unknown to him, he apparently responded enthusiastically to those who
addressed him in Spanish, while consistently ignoring those who spoke
English to him (at that time he had had virtually no contact with the
latter language).
Jakobson’s theory (1941) of the child’s sound system developing
phonemically by a series of binary splits was readily accepted by a
number of psycholinguists. Leopold (1947), in the second of his four
volumes on bilingual language acquisition, discusses the emergence of
his daughter Hildegard’s sound system during her first two years. He
looked at his data from the point of view of Jakobson’s theory of con-
PHONOLOGICAL DEVELOPMENT 57
trasts, and he found that it worked relatively well, although it did not
match in all details. One of the most interesting points of this early
piece of research is perhaps Leopold’s claim that the sound system
develops phonemically rather than phonetically (i.e. by building up
meaningful units of sound, or phonemes, rather than a succession of
sounds as they come), and that sound substitutions, far from being ran¬
dom, follow a systematic pattern.
Whereas the bilingual’s processing of the sound system follows the
same pattern as that of the monolingual speaker, the task involved is
obviously made more complex because two sound systems are in¬
volved. A larger number of features have to be recognized and
produced, and this bigger cognitive load may well lead to a later onset
of speech production or even an initial period of some confusion. Thus,
for instance. Burling (1959) observed that his son Stephen’s English
sound system developed later than his Garo phonological system (he
had more exposure to the Garo language than to English); at the age of
2:9 separation of the English and Garo vowel systems seemed to occur,
but the consonant systems never became really differentiated, Garo
consonants replacing English ones; the latter emerged properly only
when, at age 3:6, contact with Garo ceased (as the family returned
from the Garo Hills, in Assam, India, to the United States).
However, the absence of sound confusion has been remarked upon
more often than its presence. Already Ronjat (1913), one of the earliest
observers of bilingual language development this century, stated that
his son Louis realized the phonemes of German and French correctly
when he was 3:5, and that he would give the appropriate phonetic
shape to the loan words he used. More recently, Oksaar (1970), whose
son Sven acquired Swedish and Estonian (and later German as third
language), noted that there was no confusion of the sound systems. As
an example she describes her son’s mastery of the prosodic feature of
length, which is a prominent characteristic of Estonian (but not Swed¬
ish) vowels and consonants. The child had acquired the three different
degrees of length before internalizing all the segments of Estonian, and
he never used Estonian length features when speaking Swedish. Simi¬
larly, Raffler-Engel (1965) reported no confusion with respect to sound
in her son’s English and Italian, although there was a certain amount of
mixing in the morphology of both; she interpreted this dichotomy be¬
tween phonology and morphology/syntax in the child s speech as a
clear reflection of the duality of structure, a well-known basic charac¬
teristic of language: it is maintained that language is not organized
monolithically, but on two different levels - the phonetic/phonological

and the morphosyntactic.
In connection with this last point McLaughlin (1984) refers to the
theories implicit in the so-called ‘Critical Period’ hypothesis (see sec¬
tion 2.1.2). As is generally known, linguists of the Transformational
Generative school proposed the view that the child has a special capac¬
ity for language that the adult has lost. After reviewing the literature on
the hypothesis, McLaughlin concludes that there is no clear evidence
as to how long the critical period lasts or how it relates to lateralization
of language function: in other words, we cannot prove anything about
the nature of the relationship that may exist between the specialization
of the human brain, on one hand, and the ending of its special capacity
for acquiring language, on the other. With regard to language func¬
tions, many scholars would say nowadays that there is more flexibility
than was previously supposed, and that the functions relating to the
neuromuscular patterns involved in speech (i.e. the production of
speech sounds and prosodic features) may operate quite independently
from other language skills. Authors on child language acquisition also
emphasize the relative importance of social and cultural factors over
the purely biological ones.
The Critical Period hypothesis seems more convincing when applied
only to motor aspects of language acquisition. There is a quite consid¬
erable body of conclusions based on research on second language
acquisition and second/foreign language learning which shows that
younger learners are better at acquiring a native-like accent than older
ones. But there are also those who argue that, given proper instruction,
a favourable learning environment and the right psychological attitude,
older learners, too, are able to achieve native-like control of the pro¬
nunciation of a second language.
With regard to the phonological development of bilingual children,
it seems that the two systems are largely acquired separately. During
this process the child may adopt certain strategies, such as substitution
or avoidance of certain features, just as monolinguals do. Rulê-Dravi-
n,a (1965; 1967) observed her two children learning languages with
different r-sounds: Latvian, which has an apical roll /r/, and Swedish
(or, rather, a southern Sweden variety with an uvular roll /R/, similar to
the German). She found that the uvular roll was acquired before the
apical one had been properly mastered, and that the former tended to
be used in both languages. The same phenomenon occurred in my
daughter’s speech for a while (Hoffmann 1985). She initially acquired
PHONOLOGICAL DEVELOPMENT 59
Spanish and German, languages in which there is an apical and an

uvular roll respectively. By the age of three she had internalized both
phonological systems with the exception of the Spanish /r/ and some
German and Spanish consonant clusters, notably those involving the
uvular and apical roll. At the beginning the /R/ replaced the Spanish /r/
in initial position, e.g. in rojo, which she pronounced, for a time, as
[Roxo]; I was able to observe only one instance of substitution of /r/ -
by an unaspirated /t/ (in the name ‘Mary’, which she pronounced
[meti]. In syllable-final position /r/ was avoided, e.g. mi hermano was
realized as [mi eismô], so that the /r/ appeared to have been dis¬
placed by a dark schwa, which does occur in German in this position.
In consonant clusters /r/ was omitted for some time, e.g. padre [pabe]
fruta [fiita], etc., just as some monolingual Spanish children do; but in
Cristina this period of non-production was probably longer. Certainly,
it was much shorter in the case of her younger brother, so that perhaps
it may have been an individual difference, without linguistic signific¬
ance. Cristina did not add the phoneme /r/ to her repertoire until after
the equivalent clusters had been acquired in German, i.e. drei had
changed from [dai] to [dRai], Frosch from [faj] to [fR^j], etc. Another
example of simplification strategy was provided by my son, who for
some time (starting at 1:9) referred to both a frog and a duck by the
onomatopoeic ‘quack-quack’, which he pronounced [â’âk]. He re¬
placed this difficult sequence (Spanish cudc [kwak] or German quack
[kvak], i.e. a voiceless velar plosive followed by either a semivowel
[kw], or a voiced labiodental fricative [kv]) by a voiced bilabial frica¬
tive [p], a sound that occurs neither in German (anywhere) nor in
Spanish (in initial position). Compared with his sister’s general
phonetic development, his progression was rather slow, since his ac¬
quisition of sounds lasted well into his fourth year (a third language,
English, having been added to his regular exposure by then). The fea¬
tures of simplification, particularly of replacing consonant clusters by
single consonants, was very noticeable in Pascual, and also the fact
that he used sequences of sounds which did not appear (either at all or
in the positions in which he used them) in any of the languages he was
acquiring. (Both children, at the time of writing aged thirteen and ten,
pronounce their three languages without traces of the phonetic pecu¬
liarities that characterized their earlier acquisition stages.)
Apart from some initial mixing and some ‘blends’ (forms made up
of two languages, with one phonetic shape used in either language -
see Vihman and McLaughlin 1982; Grosjean 1982: 184 gives the
example ‘shot’ as a blend of French ‘chaud’ and English ‘hot’, from a

two-year-old English-French bilingual), bilingual children acquire the
language system of each language side by side and seem to develop a
feeling for their distinctiveness. This is borne out by observations that
bilingual children will borrow words from one language and use them
in the other after giving them the appropriate phonetic shape. But the
extent to which children keep the two systems apart depends, possibly,
on linguistic factors, and certainly on environmental ones. It has been
argued, e.g. by Rulê-Draviiia (1965; 1967) and Hoffmann (1985) that
interference will be the more likely the closer the languages involved
are in their phonological and morphological characteristics, intonation
features being the first to be affected. Metraux, in her 1965 study of
English-French bilingual children, comments on the importance of
playmates and friends in counteracting the influence of accent inter¬
ference. My own observations would appear to confirm this. In
children, to a greater extent than in adults, prolonged absences from a
particular speech community result in loss of fluency of intonation. But
once contact with the spoken language, especially that of other child¬
ren, has been renewed, an enriched repertoire of stress, rhythm and
intonation features can be noticed, as well as the ability to apply subtle
variations (e.g. in role-play and story telling). It stands to reason that
one must recognize the need for varied and continued language input.
If one of the languages the child is exposed to dominates considerably
over tlie other, then the subordinate language is likely to be affected by
it at the phonological level as well as all others. And if exposure to one
language ceases, it will not be long before that language disapjxiars
from the child — in the case of Burling’s son (1959) it was a matter of
only some six months; Murrell’s daughter (1966), who was bom in
Helsinki, began with three languages around her, but exposure to Fin¬
nish was irregular and it ceased at 2:1, so the little of this language that
she had acquired vanished in some four months.
3.3 Semantic and lexical development
3.3.1 General acquisition processes

The cognitive processes which underlie the development of the
meaning system are essentially dependent on the intellectual maturity
of the individual child, and they are basically the same for all children,
irrespective of whether they are acquiring one language or growing up
SEMANTIC AND LEXICAL DEVELOPMENT 61
as bilingual speakers. The task involved is to attach phonetic forms to

referents (objects, events, activities, sensations, etc.) and to abstract
concepts so that, for instance, not only all creatures with ears, a tail,
four legs, covered in fur and making certain kinds of noises are ‘dogs’,
whatever the length of the ears or tails or their vociferousness, but also
representations of them in plastic, soft material or books, just as the
family pet who is generally referred to as ‘Fido’ is a dog, too. The
young child’s first utterances resembling adult speech consist of words,
or ‘holophrases’, and they may have several meanings and functions.
‘Teddy’ may be used to refer to all kinds of cuddly toys, just as it may
be uttered with the meanings ‘Here’s a/my teddy’, ‘I want my teddy’,
‘Where’s Teddy?’ or even ‘I want to sleep’ (the teddy being associated
with sleeping because it is taken to bed).
Overextension of meaning is very common in young children. By
referring to all moving machines (lorries, cars and airplanes) as [i:jum]
Pascual at 1:9 concentrated on the characteristics of movement and
noise, disregarding features such as shape or size. The above-men¬
tioned example of [Pa’Pak] is another instance. De Villiers and De
Villiers (1979) recount a charming example of overextension: when
their son Nicholas was about one year old he learnt to call the family
dog by its name, Nunu; he then used the word for all dogs, and went
on to use it for other animals as well, and even for slippers and coats.
When he was 1:1 he had his first encounter with a black olive (in a
salad), which he referred to as Nunu, too. After some puzzlement, the
parents recognized the similarities of the features ‘round, black, shiny,
wet, cold’ with Nunu’s nose!
As the child’s vocabulary increases, (s)he learns to differentiate
meanings and acquires the necessary linguistic forms to express these
(finer and finer) distinctions. Certain items used during the one-word
stage are abandoned, notably those that are dissimilar to the adult
forms and are not kept alive as ‘family words’. Thus Pascual’s [i:jum]
developed, eventually, into ‘Auto/coche/car’, ‘Flugzeug/avion/aero-
plane’, etc. There have been reports on individual differences, and on
general trends, in the way in which children’s vocabulary develops -
for instance, by Hernandez Pina (1984), who observed various aspects
of her son Rafael’s acquisition of Spanish, paying particular attention
to semantic and syntactic development. Peters (1977) distinguished dif¬
ferent approaches to learning, which she called analytical (i.e.
beginning with the parts and building up the whole) and ‘Gestalt’ (that
is, trying to produce complete phrases and then taking them down.
therefore going from the whole to its parts). Others have emphasized
individual differences among children, e.g. Nelson (1973), who divides
them into ‘referential learners’ and ‘expressive learners’. Allowing for
certain individual differences, then, all children use overextension and
differentiation of meaning, and all abandon some earlier forms as their
lexis increases.
3.3.2 Lexical development in bilingual children

Leopold’s first volume (1939) contains a careful analysis of his
daughter s vocabulary development. Apart from describing the emer¬
gence of her German and English words, he also offers a semantic
classification, and some comments on her phonetic accuracy. He pays
special attention to those words that gradually faded from active use,
attemping to give reasons for their transience (Leopold uses ‘mor¬
tality’), such as avoidance of difficult forms, rejection of non-standard
usages, acquisition of more specific terms, and shifts from words used
in one language to items in the other, usually the more dominant one at
the time. Leopold s detailed studies have become reference works for
those interested in first language acquisition as well as researchers of
bilingualism.
There is considerable agreement among those who have reported on
bilingual language acquisition that, at the beginning, particularly the
one-\\mrd stage, the child uses words from both languages indiscrimi¬
nately.'Sometimes this has been termed ‘confusion’ or ‘initial mixing’
(see further discussion of mixing in Chapter 5.4). Grosjean (1982) also
points to the occurrence of certain forms, such as compounds and
blends, used at this stage, like ‘bitte-please’ (also reported by Leopold
and Saunders) and ‘pinichon’ (a blend from English ‘pickle’ and
French ‘comichon’). Mixed elements in early language production
have been interpreted as evidence of an undifferentiated language sys¬
tem. Volterra and Taeschner (1978) propose a three-stage model which
reflects their interpretation of mixing: in stage (1) the child operates
only one lexical and syntactic system which comprises items from both
languages; in (2) the lexicon becomes differentiated but there is con¬
tinued unification of syntax; and in (3) differentiation of the syntactic sys¬
tem appears. This model, and particularly the idea of an early unified
system, found at the time considerable support (see later in this chapter).
In fact, whether children who are acquiring two languages simulta¬
neously during infancy initially operate one system or two incipient
ones is currently a controversial issue. Evidence has been gathered

from all aspects of language acquisition as support for one view or the
other, but no firm conclusions have yet been reached. We shall return
to this question later (see section 4.2).
3.3.3 Acquisition of equivalents

In analysing the lexical development of two Italian-German bilingual
children, Volterra and Taeschner (1978) distinguish two stages. During
the first, the child has one lexical system only that includes words from
both languages. The second stage starts when (s)he is beginning to use
equivalents. In her 1983 book, which represents a longitudinal study of
the acquisition of Italian and German by her two daughters, Taeschner
maintains the claim for the validity of this distinction. In relation to
stage one she gives word lists which contain German, Italian and neu¬
tral words; the latter include onomatopoeic items, baby words and
proper names that have the same meaning in German and Italian. She
attaches much significance to the fact that very few equivalents were
present - and where they did occur, e.g. ‘da’ and ‘la’ (both meaning
‘there’), they were not recognized as such. Taeschner (1983: 27) de¬
scribes the relationship of these pairs of words in terms of hyponymy
(or partial inclusion of meaning), not of synonymy (or similarity of
meaning). With reference to the few other equivalents in the lists, she
explains that their occurrence was much more frequent in one language
than in the other. The first stage lasted for about five to six months,
and the vocabulary contained some 65-85 items.
Taeschner paid special attention to the acquisition of equivalents
during the second stage. She presents her data in the form of lists and
graphs, organizing her lexical analysis under traditional categories like
articles, nouns, adjectives, verbs, etc. In all cases there was a time in¬
terval between the acquisition of a new item and that of the equivalent,
ranging from three to eight months. For the older child this period
tended to be longer than for the younger one. The data show different
patterns in the acquisition of equivalents. Articles were acquired almost
simultaneously, but categories such as adverbs and conjunctions over¬
lapped in the temporal sequence or with little difference in time
between the acquisition in German and in Italian, whereas for nouns
and verbs the new acquisitions always preceded by several months that
of their equivalents in the other language. This is quite logical, in view
of the large number of content words (nouns, verbs and adjectives) in
any language. Taeschner suggests'; ‘It would seem that the child learns
first to use a word well in one language to refer to specific events or
objects, and only after having used it for a certain period of time be¬
gins to use its equivalent also’ (Taeschner 1983: 30).
Vila (1984) looked into the language development of three children,
bilingual in Catalan and Spanish, which he compared with that of a
reference group of three monolingual children (one in Spanish, two in
Catalan), focusing in particular on the increase in vocabulary in both
groups. His conclusion was that the language development was similar
in the two cases. As regards equivalents, according to his count they
amounted to only about 10 or 12 per cent of the total forms, and he
suggests that this can be explained by the linguistic proximity of the
two codes. However, if one counts the words he calls ‘neutral’, i.e.
very similar or identical in Catalan and Spanish, then the total comes
to some 29 per cent, which is much closer to Taeschner’s results.
The analysis of her daughters’ lexis at 2:4 and 2:10 respectively led
Taeschner to the conclusion that approximately one-third of their entire
vocabulary was made up of equivalents, while the remaining two-thirds
were new acquisitions: ‘It is in this relation that the bilingual child is
able to acquire two lexical systems simultaneously’ (1983: 33). To
generalize and attach actual figures to the acquisition process may
strike observers as a somewhat bold undertaking. Taeschner’s study
suffers from a certain methodological weakness, since it is based on
data transcribed in orthographical rather than phonetic form and it con¬
tains scarcely any indication of contextual features. Children’s early
vocalizations can be very difficult to interpret with certainty, and when
transcribing bilingual children’s output the transcriber may, quite invol¬
untarily, take important decisions about the particular form or language
the child under observation has used. We may have to wait for the
results of other longitudinal studies to see how much support these
figures receive. Equally, more comprehensive lists taken from the first
stage (lexical mixing) are needed to ascertain whether the absence of
equivalents is observable in all children or whether it is largely deter¬
mined by the linguistic input. We know that, in the case of very young
children, the roles of those who look after them (parents, childminders,
etc.) are usually quite clearly defined and, for this reason, the daily
routines involved in looking after the child are not often shared in an
equal way, which means that the subject is presented with varying lan¬
guage input and models, in contexts at least partially separated, and
this can have some influence on the acquisition of equivalents. In any
case, the absence of an equivalent can be interpreted in two ways: it

may be an item genuinely not known yet or, simply, a form not pro¬
duced; limitations in the sampling procedures are such that in many
cases one cannot be sure. The acquisition of equivalents seems to be of
relative importance only in the early development of bilingualism.
As a last point, it may be added that equivalents are rarely learnt by
analogy. Bilingual children seldom recognize cognates (e.g. ‘Papier,
papel, paper’), nor do they attempt to refer to cognate forms in one
language consciously in order to deduce the probable shape of equival¬
ent words in another. It seems that the same happens when bilingual
children begin a foreign language at school. Children, at that stage,
lack the maturity for such analytical linguistic operations.
3.3.4 The semantic load

The importance of the child being able to keep the two systems apart
has been mentioned earlier. Different contexts, environmental or
human, enable the child to distinguish the codes and to develop
strategies for language choice. This process requires a considerable
cognitive effort. In assigning words to meanings the bilingual child has
to carry out the two-fold task of allocating two labels, rather than one,
for each semantic unit. Apart from this, relating words to meanings
always includes making generalizations and abstractions, but for the
bilingual it requires, in addition, the recognition that not only do ob¬
jects, events, actions, etc. have two names, but the semantic relations
between them can be different in one language and the other. The ca¬
pacity for acquiring new words and equivalents in bilinguals is
therefore subject to individual variation: it depends on such variables
as cognitive maturity and memory, and also on interactional factors
related to the sociocultural environment of the child. The bilingual’s
linguistic capacity has to encompass the two languages. But this does
not mean that his lexicon is twice as big as that of the monolingual.
The bilingual is able to denote the same number of objects, events, etc.
as the monolingual, although he does so by dividing his linguistic
repertoire between items from both codes, with varying degrees of
overlap.
The driving forces behind language development are the communi¬
cative needs. In order to express what (s)he wants to say the child does
not have to possess a particularly rich vocabulary - an adequate one
will do. Doyle, Champagne and Segalowitz (1978) compared twenty-
two bilingual children with twenty-two monolingual ones, all aged be¬
tween 3:6 and 5:7, and they found that the monolinguals had a larger
vocabulary than the bilinguals in their dominant language. But the bil¬
inguals showed superior verbal fluency in their story-telling, and also
in the number of concepts per story expressed by each child. So,
whereas the semantic load involved in the lexical development is clear¬
ly bigger for the bilingual than for the monolingual child, the overall
resulting lexicon need not be (in fact rarely is) twice as big. Successful
communication depends less on the number of lexical items a child
possesses than the way the available ones are used. There are those
who believe that, because of their familiarity with more than one lin¬
guistic system, bilingual children are more flexible and creative in
handling language (a point that will be taken up in Chapter 7).
3.4 Development of grammar
3.4.1 Morphological development

Morphology has received less attention than syntax in the various
studies on bilingual language acquisition. Languages vary quite consid¬
erably as to the complexity of their morphology, and whereas in some
(e.g. English) the absence of morphological markers is not likely to
lead to a frequent breakdown in communication, in others (for
exarqple, the Romance languages in their verb systems) the accurate
use of inflections is very important. Many of the case studies listed in
section 2.4 involve languages of the former kind; this may explain, in
part at least, the relative scarcity of data on morphological develop¬
ment. In this section reference is made to some of the observations on
morphology made by various authors, before proceeding to discuss
syntactic development.
Leopold (1949a) noted that by the end of her second year his
daughter’s morphology was still largely undeveloped. Apart from the
possessive marker (which is the same in English and German), she
used some plural forms (mainly regular, only some irregular ones) in
English, but hardly any in German. She had no noun or verb endings
in German, and only a few adjectival endings in English derivations.
Burling’s case study (1959) is interesting because it involved a non-
European and highly inflected language. He reported that his son’s
morphological and syntactic development was already well under way
by the time the boy was two. He could use Garo suffixes indicating
DEVELOPMENT OF GRAMMAR 67
future, past, imperative, present and habitual aspects, and other ca¬
tegories, on all verbs he used, and he was beginning to use negatives,
possessives and some noun suffixes. For English Burling had recorded
only the possessive marker. But Garo was the dominant language for
the child at the time, and this must have contributed to his faster ac¬
quisition of the Garo morphology than the English one (in the latter
there are, in any case, fewer inflections). Murrell (1966) summarily
says that his daughter did not use any morpheme affixes at all before
she was 2:8. My own data show that the first markers both my children
began to use systematically, in their second year, were: the possessive
-s for German and also, incorrectly, for Spanish, although not for very
long; the morphemes for plural in nouns, adjectives and verbs in Span¬
ish; and some plural forms, mainly with nouns, in German.
Incidentally, later, when he was just three and had started to use some
English as well, Pascual went through a period during which he added
an -s (the English, or the Spanish, plural marker?) to many German
forms which were already plural, as in:
Pascual (3:1): ‘Mami, nimm die Blumen^ aus die Suppe’

(‘Mummy, take the flowers out of the soup’
- he called parsley ‘flowers’)
Pascual (3:2): ‘Enter die Decke, ihr Kinders!’

(‘Get under the blanket, children!’, as he was putting
his toy animals to bed)
It has been observed (see, for instance, McLaughlin 1984) that,

generally speaking, syntax develops before morphology except in those
languages in which a very close relationship holds between the two.
Morphology is an aspect of language acquisition that may make de¬
mands on quite complex cognitive processes. Apart from the general
structural relationships with other morphemes (one and the same mar¬
ker may signal several grammatical functions, e.g. case, gender and
number, in many German endings), there are also perceptual aspects,
as suffixes, they receive less salience in the pronunciation and may
thus not be noticed by the child. Also, languages contain a variety of
clues about grammatical categories and functions (object, agent, etc.),
so that communication can be quite successful without the use of mar¬
kers on nouns, adjectives and verbs in many instances; and if they are
not perceived as vital in every case, it is not surprising that it takes the
bilingual child a little longer to achieve this difficult learning task with
total accuracy.
Vihman (1985) states that inflectional markers in Estonian presented
considerable difficulties to her son, and that this could account for his
use of the easy English equivalents in certain cases. She concluded that
the use of English (e.g. ‘me’ and ‘mine’ for a range of Estonian equi¬
valents) could be seen in part as a strategy for putting off production of
the Estonian system until he had more time to assimilate it. Vihman’s
examples were taken from the time when Raivo was between 1:8 and
1:11. If she is correct in her assumption, it certainly shows that the
child had morphological awareness. Another example of such sensitiv¬
ity towards morphology was provided by my daughter when she was
3:1 - she used the word ‘socia’ (the feminine form of the noun ‘socio’,
member e.g. of a club) which, although correct, is not commonly
used in Spanish; she had certainly not heard it previously.
Child: ‘Êsta cerrado el club los sabados?’

(‘Is the club closed on Saturdays?’)
Eather: ‘^No, esta abierto’ (‘No, it is open’)
Child: ‘^Y por que esta abierto?’ (‘Why it is open?’)
Father: ‘Porque va mucha gente, van los socios’
(‘Because many people, the members, go there’)
Child: ‘^Y nosotros somos socios?’ (‘And we are members?’)
Father: ‘Sf!’
Ylhild: ‘^Y yo tambien?’ (‘Me, too?’)
Father: ‘SI, tu tambien’ (‘Yes, you too’)
Child: ‘Entonces, yo soy soda’ (‘In that case, I am a [female]
member’)
(Hoffmann 1985: 483)
In many respects the bilingual child follows the same route as the
monolingual. However, the use of particular morphemes is language-
specific and may be traced back to separate language input (see De
Houwer 1990). Depending on the relative importance of morphology
within a given code, the correct forms are acquired sooner or later. The
errors that do occur can mostly be explained in terms of overextension
or simplification, either developmentally from within the same lan¬
guage or, far less frequently, as a result of the child’s knowledge of the
other language. The kind of errors monolingual children make (e.g.
involving irregular verb forms, plurals or case inflections) are made by
bilinguals as well; they may, however, persist much longer in them or
even become a feature of their speech. There may be some evidence

that suggests that bilinguals have a heightened sense of morphological
awareness and that they are able to express it in a way monolinguals
cannot. But this awareness does not necessarily extend to earlier mas¬
tery of the morphological systems.
3.4.2 Syntactic development

There exists a considerable body of descriptive data on bilingual child¬
ren’s acquisition of grammar. Some of the more recent studies have
concerned themselves with the question of whether grammatical ca¬
tegories emerge at the same time and in the same way in monolingual
and bilingual children, e.g. Meisel (1984) and De Houwer (1990). By
looking at bilingual children and comparing their language develop¬
ment with the results of studies into the corresponding development in
monolingual children, Meisel aimed at gaining a better understanding
of the processes underlying language acquisition. In particular, he
wanted to find out whether the semantic-pragmatic principles that
determine much of the child’s early language development apply in the
same way to monolinguals and bilinguals, and whether pointers that
the acquisition of grammar is under way (i.e. the internalization of the
morphological and syntactic features of both languages) emerge at the
same time. He looked at the language development, between 1:0 and
4:0, of two children who were bilingual in French and German, con¬
centrating on the features of word order and case markers (the latter as
evidenced in the use of articles and the pronominal systems). His find¬
ings suggest that bilingual children acquire the same items, and in the
same sequence of acquisition, as monolinguals. As a general rule, the
simpler items are acquired before the more complex ones. He also
found that, in contrast to monolingual children, the bilinguals main¬
tained from the start more rigid patterns of word order, that the
Subject-Verb-Object order was given preference over other possible
structures, and that consistent use of the verb in second position within
the sentence constituents (which is a notable feature of German syntax)
occurred earlier in bilinguals than in monolinguals. With reference to
the use of case morphemes and other markers, he noticed that bilin¬
guals were able to express syntactic functions by morphological means
earlier than monolingual German children (Meisel explains that on this
point he could not include French examples, as no comparable material
was available.)
Other authors (e.g. Leopold r949a; Imedadze 1967; Raffler-Engel

1965; Carrow 1971; Kessler 1971; 1972) had previously come to the
conclusion that the acquisition of syntax in bilingual children follows
the same principles as in monolinguals: syntactic and morphological
features are acquired earlier or later depending on how salient and how
complex they are, relative to each other. An oft-quoted example is pro¬
vided by Mikes (1967), who reported on the acquisition of locative
markers in a Serbo-Croatian/Hungarian bilingual child. In Hungarian
the locative is expressed by noun inflection, whereas in Serbo-Croat
both noun inflection and a preposition are needed. The subject learned
to express the locative in Hungarian before he could master it in
Serbo-Croat. Mikes points out that this is a feature that is learnt quite
late in monolingual Serbo-Croat children as well, and that therefore
bilingual presentation did not appear to affect syntactic development.
If the two languages express a semantic relationship by similar
grammatical means (e.g. indicating possession by intercalating a prepo¬
sition between the object and the possessor, as in die Schuhe von
meiner Schwester and los zapatos de mi hermana), then it is likely that
they are acquired simultaneously. But it should be remembered here
that syntactic forms with a similar degree of complexity in, both lan¬
guages might not appear at the same time in the bilingual. The
acquisition will be affected by the way these forms are presented to the
child, i.e. whether, and to what extent, the adults around him make
frequQnt use of them. On the other hand, if one language uses a more
complex construction than the equivalent one in the other, the acquisi¬
tion of it may be delayed somewhat.
A substantial part of Taeschner’s (1983) book is devoted to the ana¬
lysis of the syntax of her two daughters. Her approach is based on the
linguistic theory of valency grammar, which has received a fair amount
of attention among German linguists. In valency grammar the verb is
considered to be the centre of sentence stmcture. The predicate may be
defined as the element in an utterance which holds all the other ele¬
ments together (the predicate can of course be more than just the verb,
but in the child’s constmctions there may not even be a verb form).
Taeschner considers that this type of sentence analysis is a suitable
instrument to allow her to describe her daughters’ progression from
single-word utterances to more complex, adult-like structures. She pos¬
tulates three stages through which children pass, ranging from simple
nuclear sentences in the first one to more complex constructions in
stage two, and going on to still more sophisticated structures, using
connectives to indicate relationships of cause, condition, time, etc., in

the third stage. During the first period the girls made no distinction
between the two languages, and they used words from both in their
utterances. During the more advanced stages various types of construc¬
tions developed together in Italian and German; in those cases where
differences showed up, reference to the acquisition of such items by
monolinguals could provide an explanation. Taeschner, too, like Mikes
and Vila, stresses that bilingual presentation of syntactic stmctures
does not significantly affect the pattern of their acquisition.
Annick De Houwer’s study (1990) of the language acquisition of a
Dutch-English bilingual child from age 2:7 to 3:4 contains an exten¬
sive analysis of Kate’s morphological and syntactic development. De
Houwer looks at gender, plural formation, diminutive suffixes, verb
forms, verb phrases, word order and different types of main and subor¬
dinate clauses, and then she analyses her findings by reference to
comparable studies of monolingual Dutch and English language ac¬
quisition. Her findings show that bilingual language acquisition runs
concurrently, each language forming a separate, closed system, and
very little influence is noticeable from one on the other. She also
makes the interesting observation that ‘Kate’s third birthday marks a
turning point in her linguistic development in general: in both lan¬
guages, structures start to appear that were absent before. It seems as if
the child is suddenly much more intensively occupied with the formal
aspects of language and their possibilities than before’ (De Houwer
1987: 424).
In contrast to Taeschner, she does not believe in an initial unitary
language system for any stage of language development (this point is
taken up again in section 4.2). The fact that both languages are affected
simultaneously by this heightened linguistic awareness leads De
Houwer to consider the possibility of including changes in linguistic
behaviour under the general changes for which advances in cognitive
development are responsible:
Although the actual ‘contents’ of Kate’s speech production is quite lan¬

guage-specific, there appears to be a mechanism at work here that concerns
both languages at once. Now obviously, as Kate is getting older, she is also
maturing on the cognitive level. Might it not be that general cognitive de¬
velopment here leads to changes in behaviour in general, thus including the
quite drastic changes in linguistic behaviour?
(De Houwer 1990: 308)
Taeschner’s and De Houwer’s studies differ not only in their con¬

clusions but, more fundamentally, with respect to their theoretical and
methodological approaches. Whereas De Houwer adopted a holistic
view of the child’s social, cognitive and linguistic development,
Taeschner’s description is presented in terms of separate stages.
Clearly, if one describes language acquisition one has to proceed
one step at a time. This does not mean, however, that the acquisition of
a first or a second language itself follows a linear sequence. The use of
the term ‘stages’* may be misleading as it suggests chronologically
successive steps that can be counted, which in turn implies that one
phase has to be completed before the next one starts. Language devel¬
ops gradually, sometimes by leaps and bounds, often appearing to
stagnate or even regress. Several features may develop simultaneously
and overlap, others may show a clear progression from one point to the
next. Some may regress in one language while progressing m another.
At all times the acquisition is dependent on social, psychological and
environmental factors, and these can be particularly noticeable in bi¬
lingual language development. It is generally accepted today that
unfavourable changes in the environment, withdrawal of psychological
or social support, or decrease in exposure to one of the languages (or a
combination of them) can have negative effects on the bilrngual’s lan¬
guage acquisition. He or she may, in the end, develop full competence
in only one language.
N.
Notes
1. Roger Brown (1973) uses the term ‘stages’ in a clearly defined way.
He proposes the use of a child’s mean length of utterance (MLU) as
a useful index of language development. On the basis of a sample
of 100 utterances, the MLU is calculated in terms of the number of
morphemes used. Taking MLU as the basis. Brown suggests that
there are five stages of language development, ranging from MLU
values of up to 1.75 in Stage I (which covers the one- and two-word
stages), 2.25 in Stage II and going up to 4.00 in Stage V. These
proposals have found wide acceptance in language acquisition re¬
search, and they have been used in the comparison of monolingual
as well as bilingual speech production. However, the concepts of
MLU and stages should be used charily, for at least two reasons:
differences in data collection methods and in the background of the
children studied may affect MLU and thus the comparability of
NOTES 73
stages; also, MLU as defined by Brown refers to English and cannot

automatically be transferred to other languages. Other researchers
have suggested more than five stages, e.g. Wells 1985.
Chapter 4
Aspects of bilingual competence
4.1 The bilingual’s linguistic competence
The study of child language acquisition is concerned with linguistic

aspects, such as the emergence of the phonological, morphological and
syntactic systems, the processing of linguistic items and developmental
features such as progression, e.g. in terms of semantic or grammatical
complexity. A number of other facets are also taken into account, like
the general cognitive development of the child, his/her social environ¬
ment, the language input (s)he receives and her/his interaction with
adults and other children, as these are thought to shape the resulting
linguistic competence to a significant extent (Oksaar 1983; Romaine
1984).
Some of these areas may prove to be more complex in bilingual
language development than in monolingual language acquisition, but in
principle they have the same relative value for both types of children,
and similar patterns of progression can be discerned. What will, how¬
ever, be quite different is the child’s linguistic competence. Bilingual
competence should not be seen as the sum of two separate parts, one
code and the other, but as a composite ability which may manifest
itself in the expression of competence in one language and the other
and, in addition, also in a system that combines elements of the two
and enables the speaker to use speech strategies not normally at the
disposal of the monolingual.
Grosjean makes the pertinent point that research into bilingual com¬
petence should not be conducted in terms of the bilingual’s individual
and separate languages, but it ought to take a holistic or, as he writes,
bilingual view of bilingualism. The speaker should not be considered
EARLY BILINGUALISM; ONE LANGUAGE OR TWO? 75
to be the aggregate of two complete (or perhaps incomplete) monolin-

guals: ‘Rather, he or she has a unique and specific linguistic
configuration. The coexistence and constant interaction of the two lan¬
guages has produced a different but complete language system.’
(Grosjean 1985a: 470-1).
The linguistic competence that the bilingual speaker develops in
either of her/his two languages hangs upon a number of factors (lin¬
guistic, contextual and psychological, among others). Some of them
have been discussed in Chapter 2, and others will be dealt with in
Chapter 5. The point has already been made that a bilingual speaker is
rarely equally fluent in the two languages, because the needs and uses
of each are usually quite different. But he is, in Grosjean’s (1985a)
words, a ‘fully competent speaker-hearer’ who has developed a com¬
municative competence which may make use of one language, or the
other, or the two together in the form of mixed speech.
The main concern of this chapter is to consider, first, a theoretical
issue, then some of the factors that appear to be relevant to the forma¬
tion of bilingual competence, such as language differentiation and
language choice, and lastly certain phenomena that manifest them¬
selves in the speech of bilinguals - interference, mixing and
code-switching. Some of these (e.g. mixing) have been observed to be
more noticeable in young bilinguals, whereas others (for instance,
code-switching) appear more frequently in the case of older speakers.
While the former seems to be of a more transitory nature, the latter can
become a fairly permanent feature of whole speech communities.
4.2 Early bilingualism: one language or two?
The earlier observers of bilingual language development did not ad¬

dress themselves to this question at all, but as advances were made in
the description of bilingual acquisition fundamental theoretical issues
were approached. With respect to the onset of early bilingualism, two
opposing theories emerged, approximately at the same time (the
1970s).
4.2.1 The unitary language system hypothesis

This theory holds that the bilingual child does not, initially, distinguish
between the two language systems. Instead, the child starts by using
one hybrid system, which only gradually becomes separated. Volterra
76 ASPECTS OF BILINGUAL COMPETENCE
and Taeschner (1978) proposed a three-stage model of early bilingual

development, as has already been mentioned. According to this model,
the child at first possesses one lexical system composed of lexical
items from both languages. In stage two the child distinguishes two
separate lexical codes but has only one syntactical system at her/his
disposal. Only when stage three is reached do the two linguistic codes
become entirely separate. This model seems attractively neat, and it
has found a fair number of supporters, some of whom adopt it explicit¬
ly (for example Saunders 1982a and 1988, and Amberg 1987), while
others do so implicitly, by not quoting any opposing view on syntactic
development (for instance, McLaughlin 1984) or by referring to the
‘slow separation’ or ‘final separation’ (Grosjean 1982) as a common
feature of bilingual children’s speech.
The evidence used to support this viewpoint is taken from reports of
young bilinguals’ indiscriminate use of their two languages and child¬
ren’s language mixing. Gradually diminishing mixed utterances were
observed by, among others, Imedadze (1967), Oksaar (1971), Swain
(1972: the title, significantly, is ‘Bilingualism as a first language’) and
Redlinger and Park (1980), who agreed that they reflected the child’s
initial inability to differentiate between the two codes.
4.2.2 The separate or independent development hypothesis

Aceprding to this view, bilingual children, from a very early age, are
able to differentiate their linguistic systems. This conviction was
voiced by Padilla and Liebman (1975), and it found support in Berg¬
man (1976), who postulated that in cases of bilingual acquisition each
language develops independently of the other. The idea that the bilin¬
gual child speaks with a hybrid system was rejected by Lindholm and
Padilla (1978; a and b).
In more recent years the unitary language hypothesis has come
under intense scrutiny, for instance by Meisel (1986 and 1987), De
Houwer (1990) and Genesee (1989). They argue that there is no con¬
clusive evidence to support the existence of an initial undifferentiated
language system, and they also point out certain methodological incon¬
sistencies in the three-stage model. The phenomenon of mixing is
interpreted as a sign of two as yet imperfect systems existing side by
side, rather than as evidence of one fused system.
EARLY BILINGUALISM: ONE LANGUAGE OR TWO? 77
4.2.3 An assessment of the two hypotheses

Meisel (1987) sees early mixing in bilingual speech as ‘a deficiency in
the [child’s] pragmatic competence’. On the basis of his own empirical
research, he says that very young bilingual children seem to use lan¬
guage-specific syntactic constructions even more consistently than
monolingual ones, and this leads him to claim that the former are able
to differentiate between the two languages as soon as they begin to use
syntactic means of expression (as opposed to purely semantic-prag¬
matic ways, which are common to monolingual and bilingual children).
He further proposes the separate term ‘fusion’ to refer to the ‘alleged
inability to separate the two systems’, a state he does not recognize as
existing among the children he studied. Interestingly, there is a foot¬
note in Weinreich (1968: 11), in which he refers to the work of the
linguist M. H. Roberts (1939), who also made the distinction between
‘fusion’ and ‘mixture’, the former referring to the generative process
and the latter to the established result.
Annick De Houwer’s original study bore the telling title of ‘Two at
a time’. Her thorough discussion of the mixed language hypothesis
leads her to conclude that finding positive evidence for the single-sys¬
tem stage is virtually impossible, and this is borne out by the findings
of her own case-study. She collected two sets of speech data from a
young Dutch-English bilingual. One contained utterances only in
Dutch and the other only in English. The corpus enabled her to address
the question of whether, and to what extent, the girl’s two languages
were developing separately. The evidence that she found led her to the
following conclusion:
In all aspects of language use investigated that provided unambiguous op¬

portunities for discovering either the presence or the absence of
inter-linguistic interaction, we were able to show that Kate’s developing
morphosyntactic knowledge of Dutch could not function as a basis for her
speech production in English, or vice versa. Instead, Kate mostly used
Dutch morphosyntactic devices when producing utterances with only
Dutch lexical items, and English morphosyntactic devices when producing
utterances with only English lexical items. Furthermore, not only were the
morphosyntactic devices themselves usually relatable to only one language,
they were also used in a language-specific manner. Therefore, I believe, it
has been convincingly demonstrated that the Separate Development Hypo¬
thesis accurately describes a major part of Kate’s bilingual acquisition
process.
(De Houwer 1990: 338-9)
As we can see, she based her conclusion on cases that ‘provided

unambiguous opportunities’, i.e. so different in English and Dutch that
no possibility existed for the subject to use morphosyntactic devices
belonging to one language as the ‘basis for her speech production’ in
the other. But what about the other cases? It has to be borne in mind
that Dutch and English are fairly closely related languages, i.e. there
are many similarities between them, a fact that could perhaps account
for a large part of Kate’s linguistic development. One is therefore en¬
titled to doubt whether anything can be proved decisively in this
matter.
Genesee’s (1989) contribution to the discussion is made in a review
article. He re-examines a number of studies of early bilingual develop¬
ment in which mixing had been interpreted as evidence for a unitary
system. Again, the unitary language hypothesis is queried for empirical
reasons. In order to uphold the theory, Genesee suggests, ‘one would
need to establish that, all things being equal, bilingual children use
items from both languages indiscriminately in all contexts of com¬
munication.’ (Genesee: 1989).
The opposite theory, in contrast, would need to prove that bilingual
children ‘use items from their two languages differentially as a func¬
tion of context’. Thus, in a situation where the young child’s weaker
language was being used (around him and by him) one would expect a
higher proportion of items in the weaker language to occur than in
contexts where the stronger language was employed. As Genesee
points out, most studies supporting the initial hybrid language thesis do
not present or analyse their data by context. And he adds: ‘Therefore, it
is impossible to determine whether the children are using the repertoire
of language items they have acquired to that point in a differentiated
way.’ (Genesee: 1989).
This author refers also to research on the perceptual abilities of very
young children, saying that they seem to be capable of fine discrimina¬
tions, such as noticing phonetic contrasts, and to possess all the
necessary prerequisites for speech perception. The assumption is, pre¬
sumably, that it does not make much sense to suggest that bilingual
children are operating only one system when, at the point of their de¬
velopment in which they begin to utter single words, they are already
able to perceive two different phonological systems.
It is difficult to see how one can confidently pronounce in favour of
one of the two models, although it does appear that the balance is
beginning to tip in favour of the separate language development hypo-
LANGUAGE AWARENESS AND DIFFERENTIATION 79
thesis. A major obstacle is that we do not as yet have a proper basis for
deciding to what extent linguistic knowledge in the very early stages is
language-specific, or up to what point we can legitimately postulate a
language competence from the child’s first utterances. Also, if one sees
children’s early linguistic competence in terms of several systems
rather than just one, it becomes possible to argue that some systems are
separate and others fused. In any case, the tools available for descrip¬
tion are crude. It can also be misleading to speak of the child’s ‘lexis’
or ‘syntax’: one would have to define very carefully what precisely
constitutes a lexical item, and allow a more general view of syntax
when looking at children’s one- and two-word utterances, as these can
have a variety of meanings and functions.
4.3 Language awareness and differentiation
If one believes that children acquiring two languages in early infancy

go through a stage during which they operate one system only, one
needs to determine exactly when bilingualism begins. Different
answers are possible, depending on one’s approach. Volterra and
Taeschner (1978) argue that only when the child ceases to classify
speakers in terms of their language can (s)he be claimed to be truly
bilingual, i.e. only when the systems themselves are acknowledged as
possible ways of communicating. A more linguistic, rather than cogni¬
tive-behavioural, view is to say that the use of synonyms, or
equivalents, is the clearest sign that the child is bilingual. The emer¬
gence of bilingualism and language differentiation are seen as closely
related issues. Areas of particular interest concern questions such as
whether there is any kind of systematicity in early mixed utterances,
how the child overextends meanings, and when and how he eventually
learns to separate the two language systems he is in contact with. The
underlying belief is that only when it is evident that the child is begin¬
ning to use two languages and to keep them apart is there proof that he
is on his way to becoming truly bilingual. On the other hand, if one
subscribes to the view that bilingual children operate two distinct lan¬
guage systems right from the beginning (in terms of perception and,
even if imperfectly, production), the question of the onset of bilingual¬
ism becomes irrelevant. Similarly, terms such as ‘separation’ or
‘differentiation’ of languages make no sense, since they imply that
prior to differentiation there was a mixed form.
We can now consider some observations that have been made about
the bilingual child’s emerging language awareness and the possible
manifestations of separate language development. The use of the term
‘differentiation’ is not meant to indicate a particular theoretical stand¬
point.
4.3.1 Manifestations of language awareness

A look at some of the case studies will show very considerable vari¬
ation in the ages at which the children observed are reported to have
become aware of operating in two different systems. This is not sur¬
prising, as it may well be the case that one child is conscious of the
presence of two languages in her environment long before commenting
on it or showing any outward signs, while another child may remark
on his observations much sooner or manifest his awareness by indirect
means. Linguist parents of bilingual children have provided examples
of amazingly early, and very perceptive, language awareness (e.g. Ron-
jat 1913; Saunders 1982a; Fantini 1985; Clyne 1988). However, at
least some of the - often highly amusing - remarks by very young
children on their own and other people’s language use must be taken
as a reflection of the emphasis and keen interest in all matters linguis¬
tic that they encounter in their environment.
There are many studies, older as well as newer ones, of bilingual
childj-en which remark on the apparently indiscriminate use of both
languages at the initial stage of the child’s speech production and then
report that the early mixing subsided once the children involved re¬
alized that there were two languages in their environment. This was
often, in essence, a reference to the realization by the children that
people around them (often the parents) used different languages, which
was acknowledged by some kind of comment. Thus Ronjat (1913) de¬
scribes how his son Louis went through a period of testing the names,
phonetic shapes and linguistic allocations of objects (Ronjat refers to
this as the ‘temps d’essai’). By using words with French or German
pronunciation the boy would ask for clarification and assign them to
the mother’s or the father’s repertoire (e.g. ‘le easier maman’ or ‘le
easier papa’). This was reported as happening before the child’s second
birthday. Leopold (1949b) says that his daughters Hildegard and Karla
were conscious of using two different systems early in their third year,
and that there was active separation of the codes towards the end of
that year, i.e. each language clearly developing separately.
My daughter Cristina showed signs of distinguishing the two sys¬

tems around her second birthday. Addressing each person around her in
the correct language, she would sometimes comment on the fact that
her parents gave objects different names, and she often checked
whether she had got the two labels right by saying, for example:
Cristina: (Looking at her mother, while banging the table)

‘Papa mesa’ (= ‘Papa sagt “mesa”’).
Or, after I had asked her (in German) to pick up a pencil:
Mother: ‘Nina, heb den Bleistift auf’
she would ask her father:
Cristina: ‘Papa lapiz?’ (= ‘Papa, ^tii dices “lapiz”?’)

(‘Daddy, you call it “pencil”?’)
Kessler mentions the first occasion, at age 3:6, when Lita, a Span-
ish-English bilingual in Texas, actually named her two languages (the
exchange occurred after her mother asked Lita to give several exam¬
ples of Spanish and English words):
Mother: ‘^Tu sabes hablar mucho o poquito ingles?

(‘Can you speak a lot or a little bit of English?’)
Lita: ‘Mucho ingles.’ (‘A lot of English’)
Mother: ‘^Y tu sabes hablar mucho o poquito espahol?’

(‘And can you speak a lot or a little bit of Spanish?’)
Lita: ‘Mucho espanol’. (‘A lot of Spanish’)
Mother: ‘Y oye, Lita, â ti te gusta hablar el ingles, o el espanol, o

los dos?’ (‘And tell me, Lita, do you like to speak English,
or to speak Spanish, or both?’)
Lita: ‘Dos’ (holds up two fingers) ‘That’s ingles’ (points to one

finger) ‘And that’s espanol!’ (points to another finger)
(Kessler 1984: 42)
An earlier example from the same child shows clearly that this little
girl was aware of two distinct codes at age 2:3:
Mother: ‘ ^Que quieres, Lita, quieres leche?’

(‘What do you want, Lita, do you want some milk?’)
Lita: ‘No leche’.
Mother: ‘^Quieres agua?’ (‘Do you want some water?’)
Lita: ‘No agua’.
Mother: ‘^Quieres jugo?’ (‘Do you want some juice?’)
Lita: ‘No jugo. Candy, mami’.

Mother: ‘No, candy no, Lita’.
Lita: ‘Candy!’ (with great emphasis, followed by a long pause)

. . . ‘Dame dulce, please’ (‘Give me candy, please’)
(Kessler 1984: 41).
Carolyn Kessler adds the comment: ‘The language switch from

English to Spanish in asking for candy took place in a very slow, de¬
liberate manner, evidently with the expectation that the situation to
bring about the realization of candy called for Spanish.’ (Kessler
1984: 41).
These examples are interesting from two points of view: they show
clearly the child’s awareness of the two separate systems, and their
appropriateness for use with particular people, English- and Spanish-
speaking; and they appear to indicate that she is unaware of mixing
them when she says: ‘dos . . . that’s ingles. . . and that’s espanol’ and
later: ‘Dame dulce, please.’
At the age of 2:6 Cristina was able to express her awareness of the
two languages by using labels like ‘so wie Mami’ and ‘como dice
papa’ (Hoffmann 1978; 1985). It is quite normal for small children not
to use abstract labels such as ‘English’, ‘German’, etc., as the concept
of language, and correspondingly ‘of the English language’ etc. are not
present in their minds until a much later stage in their cognitive devel¬
opment. The paraphrases they use often reflect the close association,
for them, between language and person(s); for instance, Pascual used
‘wie Mami’ and ‘so wie wir, Mami?’ to mean ‘[in] German’, and
‘como dice papa’ for ‘[in] Spanish’, which he extended to ‘wie Ian’
and ‘como dice Ian’ (his friend next door) for the meaning ‘[in] Eng¬
lish’. Another type of paraphrase made use of the different ways of
saying ‘hello’, i.e. as a linguistic label rather than as a form of address;
LANGUAGE AWARENESS AND DIEFERENTIATION 83
Pascual (3:7): (telling me off for speaking English to a Spanish-speak¬

ing girl)
‘Mami, nicht so mit Carmen sprechen!’
(‘Mami, don’t speak to Carmen like that!’)
Mother: ‘Wie denn?’
(‘What should I say, then?’)
Pascual: ‘So, “hola ^como estas?”, nicht “hello”!’
Only after his fourth birthday did he start to use the labels ‘Span¬
ish’, ‘German’ and ‘English’.
Cristina’s active language separation became consistent from the
two-word stage onwards, but for Pascual separation became evident
later; he was generally slower to start developing his languages.
Arnberg gives some examples of a two-and-a-half-year-old who did
not yet appear to keep apart his two linguistic systems, English and
Swedish:
‘Titta, bunny’’ (‘Look, bunnie’)

‘En block’ (‘A block’)
‘En piggies till’ (‘A piggies more’) [i.e. ‘one more pig’]
‘Har ar budda’ (butter) (‘Here is butter’)
‘Horsie sova’ (‘Horsie sleep’)
‘Ar det ducksT (‘Is it ducks?’)
(Arnberg 1987: 69)
Arnberg uses these and other sentences as examples of the child’s

attempt to combine his two languages into a single system. They can,
of course, also be cited to support the opposite view, although it has to
be recognized that the samples are very limited and there is no indica¬
tion of the linguistic context in which they were uttered, or of what the
input was (linguistic and otherwise). The child was using words from
both his languages, but we can observe a certain pattern: the grammati¬
cal items (‘en’, ‘det’, ‘en till’) and the verb forms (‘ar’, ‘titta’) are in
Swedish, whereas most of the content words (nouns) are in English. It
looks as if the child was signalling his knowledge of two systems but,
at this early stage, he still needed to use lexical items (English nouns)
in one language to fill in gaps for their equivalents in the other. Per¬
haps he just did not know them yet, possibly he still associated the
animals he was commenting upon with a certain person who spoke
English, maybe he knew the item in Swedish but could not recall it at
the moment of speaking. We sliall come back (in section 5.4) to the
subject of mixing.
So far we have mentioned the naming of the various languages and
also active separation, i.e. the use of separate languages in appropriate
contexts. There are other signals that the child can give which indicate
that he is aware of using two codes. He can use spontaneous transla¬
tion to recount to one parent what the other has said, unaware of the
fact that the parents may understand each other’s language (see, for
instance, Imedadze 1967; Swain and Wesche 1975; Hoffmann 1985;
Fantini 1985). Ruth Metraux (1965) obseiwed that some of the children
she studied would not react when asked something in one language by
a person they normally associated with the other. This seems to indi¬
cate that the child gets himself into one language mode when
communicating with one person or set of people, and into another with
a different (group of) person(s) (see Grosjean 1985a on this point). To
what extent this shitting or switching is a conscious operation we do
not know. It has often been noted that children dislike their parents
addressing them (or even others) in the ‘wrong’ language (if the one-
parent-one-language rule has been closely adhered to in their
upbringing). The instance of Pascual protesting at me for using English
to the Spanish girl, mentioned above, is one example. Fantini quotes an
instance of both his children’s intolerance of his breach of the language
rule when speaking to his wife in English, rather than Spanish;
\.
Carla: ‘jNo hables en ingles a mama!’ (‘Don’t speak English to

Mummy’)
Father: ‘/.Porque?’
Carla: ‘Porque mama no le gusta hablar en . . . ep. . . ’ (Looks

to Mario tor support) (‘Because Mummy doesn’t like
speaking in . . . Sp. . . ')
Mario (to everyone): ‘iEspanol! A mama le gusta hablar en es-

pahoF (‘Spanish! Mummy likes to speak in Spanish!’)
(Fantini DS.s; 76)
Arnlx'rg has an example of the same two-and-a-half-year-old cited

earlier reacting to the mixed speech of his father:
Child (to English-speaking mother): ‘Jag vill ha en SPOON’

Swedish-speaking father (to child): ‘Vill du ha en SPOON?’
Child (upset): ‘Nej, sked!’

(Arnberg 1987: 70)
We do not know whether he was upset simply because the Swedish

father used a word in English, or whether he thought that the father
was mocking him for using the wrong word - whatever the reason, the
example illustrates the child’s awareness of the two separate systems in
his environment.
43.2 How does the child keep the languages apart?

We can only speculate about how the child becomes aware of the ex¬
istence of the two languages. There have been attempts to design tests
aimed at bringing out evidence of some kind of mental device regulat¬
ing the ‘on’ and ‘off’ switching of the bilingual’s languages. Penfield
and Roberts (1959) first proposed a model that came to be called the
‘single-switch theory’ because it suggested the presence of such a
switching mechanism. Its shortcoming was that it could not adequately
account for some aspects of the performance of bilinguals, e.g. in
Stroop testing.^
A more sophisticated theory was formulated by Macnamara (1967),
the so-called ‘two-switch model’, which hypothesized an input and an
output switch. The speaker was suppo.sed to be in control of the latter,
but not of the former. Various tests were administered to bilinguals to
show that switching from one language to another takes time; for
example, the subjects had to name numerals in one language and then
in another, or to read monolingual passages and bilingual ones, where
the sentences contained words from both languages mixed at random. In
all cases, the clear conclusion was that bilingual performance takes longer.
The validity of both models has since been questioned, as the prob¬
lem of explaining the neutral status of a switching device remains
unsolved. In any case, research has now shifted towards seeking an
explanation for code-switching, a frequent and naturally occurring phe¬
nomenon in bilingual spêech (see Chapter 5). It seems unlikely that the
brain of the individual should have any kind of special mechanism to
separate the languages as, after all, the monolingual, too, keeps a num¬
ber of language registers and styles apart. His language production,
like that of the bilingual, involves selecting linguistic items. And both
monolinguals and bilinguals make wrong choices.
If one does not try to look for neurolinguistic devices, there are a
number of contextual factors that suggest possible explanations for de¬
veloping language awareness. Among these are the following:
(1) The child’s increased knowledge of the languages could help her
to differentiate them.
(2) More specifically, the bilingual’s greater familiarity with the
sounds of each language may facilitate recognition of the two
codes.
(3) The accumulation of social experience (i.e. the knowledge of the
contexts in which linguistic forms appear) will provide the child
with valuable information and thereby increase awareness of the
separate systems.
(4) The improvement of her linguistic sensitivity towards adult
standards, as well as her wish to approximate to these models of
linguistic behaviour, may also be influential.
(5) The child wishes to be understood by monolinguals, and this can
often force her to seek the correct word, or to explain what she
means in a linguistically more refined way than if she were
speaking to another bilingual.
4.3.3 The importance of context

The most influential factor in developing bilingual competence is prob¬
ably. the one mentioned under (3) above. Context, particularly human
context, is perhaps of paramount importance in deciding the status and
role of each language in the child’s mind and in helping him to acquire
the appropriate linguistic forms. The advantages of following the one-
person-one-language rule have been pointed out above. There is
evidence that children who have acquired their languages from differ¬
ent persons show less language mixing than those who have acquired
them in ‘fused’ contexts. The assumption is that, in the early stages,
and in order to avoid confusion and disorder, the bilingual child needs
to use the person as a reference point when organizing his language
behaviour, i.e. the bilingual normally chooses the linguistic code ac¬
cording to the language of the person with whom (s)he is interacting.
Naturally, the context does not have to be ‘a person’, it can be the
physical or situational environment as well. In many cases bilingual
children grow up to speak two languages by being exposed to one in
the home and another one outside, so the contexts of use do not often
merge. Another form of separation of context may be a combination of

the two possibilities, as for example with the family who decide that
they will use the father’s language when they are all together but that
of the mother in the absence of the former. A separate context for each
language is considered to be important for the establishment of biling¬
ualism, as it helps the child to understand that the two codes are
possible and distinct means of communication, existing in their own
right, rather than being attributes of certain people. This process takes
some time to be completed. I have a record of a situation involving
Pascual, when he was 1:11, which shows that he had not made that
abstraction yet. One evening he discovered a full moon shining
through the window. He was fascinated, and when I told him ‘Das ist
der Mond’ he repeated ‘Mond, Mond’ several times and went, excited¬
ly, to the next room to point out the moon to his father, from whom he
learnt to say ‘la luna’. The next day a Spanish friend of ours, Pascual
and I were sitting at the table when Pascual again discovered the moon
outside, and the following dialogue ensued;
Pascual (1:11): ‘Mira, [pija] (Pilar), Mond’

Pilar: ‘Si, la luna’
Pascual (correcting her): ‘Mond! Mond!’
Pilar; ‘Sf, la luna’
At this he got visibly upset, insisting on calling the moon ‘Mond’;

he would not accept my explanation that Pilar used other words like
‘sf and ‘nino’ where I said ‘ja’ and ‘Kind’, for example. Perhaps the
two contexts, the human and the physical, were too close for him to
separate them as easily as he had done the night before; the acquisition
of the two nouns would probably have been easier for the child if the
separation of the human context had been accompanied, all the time,
by physical separation.
The particular way and the specific context in which the language is
presented to the bilingual child are seen to be influential. The advice
given to parents who want to bring up their children bilingually is
usually that they should follow the one-person-one-language rule or
that they ought to keep to one language at home and to the other out¬
side it. It seems reasonable to expect that a separation of context and
an input of one language at a time will facilitate the child s learning
task. However, there is no empirical evidence to show that other meth¬
ods have detrimental effects, such as delayed language production. But
studies of early bilingualism do suggest that mixed input is more likely

to cause mixed speech. And there is a popular (but not well-
researched) belief that those who consistently mix languages are not
capable of keeping them apart, i.e. that they somehow have a language
deficiency.
4.4 Language choice in children
As the young bilingual child develops her social and linguistic skills,
she is likely to meet a larger number of people from different linguistic
backgrounds with whom to interact. Which language will she choose
to use? According to what criteria will she select a linguistic code?
The choices available to her can be summarized in the following way.
monolingual in monolingual in bilingual in

Lang. A Lang. B Lang. A and Lang. B
(will choose) (will choose) (will choose)
She is likely to make a straightforward choice (between Lang. A

and Lang. B) when selecting the language she is going to use to mono¬
lingual speakers. But what makes her decide to go for Lang. A or
Lang. B when talking to a bilingual? And will she stick consistently to
that choice once she has made it?
4.4.1 Language choice in monolinguals and bilinguals

Monohnguals also have to make choices when selecting certain styles
and registers, and when changing from one style to another. Contribu¬
tions by a number of linguists (e.g. Ervin-Tripp 1964, 1968; Gumperz
LANGUAGE CHOICE IN CHILDREN 89
1964; Fishman 1965a and 1971 - the latter in an article with the infor¬
mative title ‘Who speaks what language to whom and when?’) have
elicited a number of factors that can be considered to account for shift¬
ing from one variety to another:
(1) the setting, in terms of time and place, and the situation (‘do¬
main’ is another word used), e.g. whether in the family, at a
party or at the work-place;
(2) the participants in the interaction, i.e. features relating to their
age, sex, socio-economic status, occupation, etc.;
(3) the topic of conversation;
(4) the function of the interaction, which can be to greet, apologize,
exchange information, etc.
The same four categories of variables will decide, at any rate in
general terms, the bilingual’s language choice. Since two languages are
involved, the situation may become more complex, as the switches will
involve the selection of different varieties of Lang. A or Lang. B and,
in addition, switches from A to B and back from B to A.
4.4.2 Determinants of language choice in bilingual children

Essentially, the factors influencing language choice are the same for
adults and children. However, they may not operate in quite the same
way for each group. In what follows, a brief account is given of some
of the determinants of language selection discussed in the literature on
child bilingualism.
(1) The person(s) engaged in the speech event This is probably the
most important variable for the child’s language choice. In studies of
bilingual children there are numerous references to children who ad¬
dress one parent in one language and the other parent in the second. As
the child grows, he may extend the pattern so that, for a while, all men
are spoken to in the father’s tongue and all women in that of his
mother. Sometimes children may choose a language on the basis of a
person’s looks: Burling (1959), for instance, reported on his son talk¬
ing in Garo to an Indian-looking stranger he met on a plane (although
he had not used Garo for some time); and in Fantini (1985) there are
several examples of Mario associating all persons with dark complex-
ion/hair with the use of Spanish - but assigning English to black
people. A person’s role with respect to the child may be conceived of
as decisive, e.g. nursemaids, babysitters, teachers who become associ¬

ated with a particular language.
There is anecdotal reference, as well as empirical evidence, from
those who have collected data on young children’s language selection,
that children tend to follow a consistent pattern. Most of a bilingual’s
utterances to a monolingual speaker are in the language of that
speaker, and incidences of mixing may be remarkably low. There is
similar evidence that bilingual children are more inclined to address
other bilingual speakers in either language, just as they are prone to
produce more mixed utterances in their speech with bilingual interlocu¬
tors.
Children may also judge on the basis of the linguistic features ap¬
parent in the person, i.e. whether the speaker has a native accent or
not. An instance of this is that my children, when answering the tele¬
phone, have always switched to Spanish or German when they
detected a Spanish/German accent in the caller’s Enghsh. Fantini
(1985; 62) offers an interesting example of Mario’s reaction to the lan¬
guage of several Mexican-American classmates who frequently
switched between English and Spanish; Mario used English only when
speaking to them (the kindergarten was in Texas), and when asked by
his father: ‘Mario, ^por que no hablas en espahol con esos nenes?’
(‘Mario, how come you don’t speak Spanish with those children?’), the
five-year-old boy replied: ‘Ah, esos nenes no hablan muy bien’ (‘Oh,
thosb- kids don t speak very well’). Presumably, he was referring to
their Spanish as judged by the native-speaker standards he was familiar
with.
(2) The settingiplacelsituation This last example shows that another

determinant, that of ‘kindergarten playground’, may also have in¬
fluenced Mario’s choice of English. Pascual, as a two-year-old, would
greet any person in the street with ‘Hello!’ or ‘Hello man!’ without
having been spoken to by them first. But within the home other factors
were of greater importance than ‘setting’ in choosing the language in
whieh to address visiting strangers.
On a visit to her grandmother in Denmark, Cristina (shortly before
her third birthday) addressed people who spoke Danish to her in Eng¬
lish, which at the time was very much her weakest language, rather
than in German, which was spoken to her by her grandmother and her
mother. It seems that she made her language choice on the basis of
setting, associating English with ‘outside the home’. It is also possible
LANGUAGE CHOICE IN CHILDREN 91
that linguistic features played a role, i.e. she may have decided that,
since Danish was an unknown system, it somehow called for the
choice of her least well-known language.
Harrison and Piette (1980) looked at the factors determining the lan¬
guage choices that bilingual Welsh—English mothers made while
raising their children. They were able to show that most of the children
were linguistically as their mothers had intended, and that their lan¬
guage choices had often been determined by the mothers. The latter s
decisions as to whether to bring up their children monolingually or
bilingually had depended on:
(a) socio-economic considerations: mothers at either end of the so¬

cial scale were more likely to go for bilingual upbringing than
those in the middle;
(b) the chances of social interaction outside the nuclear family: for
example, the availability of Welsh-English bilingual playgroups
and schools, and contact with other bilinguals;
(c) the influence of social interaction inside the family, such as the
father’s encouragement or lack of support for a bilingual up¬
bringing; and
(d) socio-psychological factors, as for instance the mother s attitude
towards Welsh and child-rearing, or the use of a specific lan¬
guage generally.
The authors conclude that children pick up subtle cues from their envi¬
ronment, and that the mothers’ attitudes and language choices will
ultimately determine their own.
(3) The function or purpose of the interaction Harrison and Piette also
mention person and place as possible determinants of young bilinguals’
language selection. Their main argument, however, is that bilingual
children ‘use language as a lever to the world’ (1980: 218), just as
monolinguals do (see Halliday 1975; Bruner 1975). They give exam¬
ples from a child who switched from one language to another in the
hope that the person she was with would comply with her demand for
more playing, and from another subject who, while being in the com¬
pany of Welsh and English speakers who were discussing whether or
not he ought to go to bed, kept switching to Welsh because he felt that
there was a better chance that the Welsh speakers would let him stay
up. The example from Kessler (1984) given in section 4.3.1. above can
be seen in the same light. Lita consciously chose the Spanish word for
candy , after having used it in English to her Spanish-speaking

mother, in the hope of getting some.
A particular language may be chosen in order either to exclude or to
include other people present. Some parents report that their offspring
do not address them in the usual language when they pick them up
from school, for example, or when a monolingual friend is present.
Children may also decide to select a particular language, in preference
to the other, in order to re-tell a story or a joke, or to cite someone
else, or when involved in role play or quoting or making puns, or when
providing a ‘translation service’ for somebody. In all these instances
(and many others are possible), the choice is made with the intention
that the message should convey a particular meaning or have a certain
effect.
(4) Topic Only as the child gets older does this variable acquire im¬
portance. Once children are of school age and start having a host of
new experiences (special school activities, academic subjects, sports,
hobbies, TV programmes, etc.), their language selection will become
strongly affected by the topic of the conversation or language use.
(5) Linguistic proficiency The degree of proficiency which the child (or
adult) has in both languages may constitute a mainly linguistic or psy-
cholinguistic reason for language selection. The speaker may prefer to
use the language he feels more confident in because of the frustrating
expefience of not being able to say quickly and with ease what is im¬
portant at a particular moment. In this kind of situation all the other
determinants can be overridden by the one factor of linguistic availa¬
bility; and the less balanced the bilingual is in the respective
competences, the more likely it will be that this factor will come into
play. In relation to older bilinguals one would speak of language do¬
minance. With reference to the very young child, whose language
development is still an on-going process, Dodson (1981) introduces the
concept of the preferred language.
Language development is intimately linked with the child’s increas¬
ing awareness of the language of her/his environment and his/her skill
m using it. Bilingual language acquisition involves developing an
awareness of two distinct systems, acquiring their features and learning
to keep them apart. Becoming bilingual implies making choices be¬
tween two languages, following rules that are laid down by the
environment or that the individual has decided upon by himself. In
NOTES 93
young children such choices tend to be affected primarily by social and

emotional factors (as mentioned under (1) and (2) above), but as they
grow older more subtle factors (linguistic, topical and psychological)
become more influential. One should point out, though, that the deter¬
minants outlined here are all interrelated to varying degrees.
Notes
1. The bilingual version of the Stroop procedure consisted of the elici¬
tation of responses in one language while receiving a visual
stimulus in the other. For instance, the English-German bilingual
would be shown the word schwarz (= ‘black’) written in yellow ink,
and the colour of the word had to be named in English; the correct
answer therefore would be ‘yellow’. The monolingual version em¬
ployed words from one language only, i.e. black written in yellow
ink would be shown. It is known that both monolinguals and biling¬
uals are slowed down in the naming of the colour, which appears to
indicate that both systems are ‘on’ at the same time - as well as
confirming the generally observed ability of bilinguals to listen in
one language while speaking in the other, which finds its perfection
in simultaneous interpreting.
2. As many bilinguals do not have the same type of competence in

both of their languages, their stylistic repertoire in each language
may differ. This is often noticeable in children who, for instance,
have acquired one code from one side of their family and have not
learnt to use the more formal styles requiring specific morphological
or lexical forms in that language, whereas in their other language
they are able to use ‘polite expressions’ and certain forms of ad¬
dress for specific purposes.
Chapter 5
Features of bilingual speech
There are people (monolinguals themselves, usually) who advise par¬

ents against raising their children bilingually. They argue that the child
may become confused, or his intelligence affected in some way, or his
speech impaired (by stuttering), and above all they warn of the danger
that he may end up not speaking any language ‘properly’. And they
believe this last possibility can easily be proved by pointing to ‘irregu¬
larities’ present in the bilingual child’s speech (or, for that matter, that
of bilingual adults).
It is true that, when listening to bilinguals talking to each other, we
can notice certain features which are absent from monolingual speech,
such as a ‘foreign accent’, the incorporation (wholesale or in an
adapted form) of words or expressions from the other language, and
sudden switches from one code to the other, the latter perhaps occur¬
ring more than once within the same utterance.
If one holds the view that a bilingual is, linguistically speaking, a
composite containing two separate parts (or codes), then the basis for
assessing his language competence will be monolingual standards of
proficiency in the use of the two systems. In this case, notions of
purity of language will be rated more highly than those of communica¬
tive competence. But if one looks at the bilingual speaker holistically,
as Grosjean (1985a) suggests, i.e. as someone possessing bilingual
competence that can manifest itself in various ways, then the picture
can be quite different. Phenomena such as interference, mixing and
switching become the, subject of analysis, helping us to discover pat¬
terns and relationships with other features of speech. A non-native
accent and the choice of a wrong word (perhaps a loan translation from
the other language) are more likely to be detected in bilinguals when
INTERFERENCE 95
they are fatigued or excited. Children tend to mix more if they are
frequently exposed to mixed speech. And both bilingual children and
adults appear to mix and switch more when they are in each other’s
company than when talking to monolinguals - indeed, they may well
have their reasons for such linguistic behaviour, such as signalling
group identity to outsiders or solidarity to other group members, or
expressing a shared experience. Certain aspects of bilingual speech will
now be dealt with in a little more detail.
5.1 Interference
Most of the authors who have observed bilingual speech have re¬
marked in one way or another on the transfer of elements from one
language to the other. The amount of such elements noticed and the
importance attached to them differs greatly, however. In the older lit¬
erature all instances of transfer tended to be subsumed under the
heading of ‘interference’, and the definitions that were put forward did
reflect a slightly negative assessment of the phenomenon. In 1953,
when he first published his book Languages in Contact, Weinreich de¬
cided to call interference ‘those instances of deviation from the norms
of either language which occur in the speech of bilinguals as a result of
their familiarity with more than one language, i.e. as a result of lan¬
guage contact.’ (1968: 1). In the definition given by Mackey there is,
however, no reference to ‘norm’ or ‘deviation : Interference is the use
of features belonging to one language while speaking or writing an¬
other’ (Mackey 1970: 569).
Before going on to examine the various types of interference - pho¬
nological, lexical, grammatical and cultural - Mackey emphasizes the
distinction between interference and borrowing: the former is an in¬
stance of ‘parole’, while the latter is one of ‘langue’. Interference, he
argues, may vary a good deal according to a number of psychological,
situational and discourse factors.
In more recent studies, features of bilingual speech have been dealt
with under the separate headings of interference, borrowing, mixing
and code-switching, which reflects the various characteristics that have
been discerned. But, as so often happens in the fields of linguistics,
there are no clear-cut distinctions or commonly agreed approaches to
analysis or description, and the definitions one comes across may, at
times, seem contradictory. In other cases some of the descriptions may
96 FEATURES OF BILINGUAL SPEECH
overlap, so that the task of separating, for example, discussions of mix¬

ing from those of switching is not as easy as one would like it to be.
Each definition offered expresses the researcher’s views on and ap¬
proaches to a particular issue.
Grosjean (1982 and 1985a), like Mackey, prefers a neutral defini¬
tion of interference: The involuntary influence of one language on the
other’ (1982: 299) - which distinguishes it from borrowing and code¬
switching, understood to be less involuntary. The latter are quite
noticeable features of the speech of bilinguals when addressing each
other, whereas interference is also prominent in the speech of bilin¬
guals when addressing monolinguals. According to Grosjean, a
bilingual, when speaking to a monolingual, realizes that code-switch¬
ing and borrowing would impede ease of communication and he
therefore avoids them. Three points should, however, be taken into ac¬
count here: first, that speakers do not always consciously control their
speech so as to avoid borrowing and code-switching; second, that situ¬
ational and emotional factors can affect interference so that, for
instance, the Spanish—English bilingual who is normally proficient at
producing the English [r], in initial position, and omitting it in postvo-
calic position, will shout ‘Be careful!’ with a heavily rolled alveolar [r]
after the first vowel when agitated; and lastly, that the distinctions be¬
tween the four phenomena are less clear-cut in young bilinguals. With
respect to the very young bilingual still undergoing the process of lan¬
guage, acquisition, it makes little sense to talk about interference at all
(particularly if one assumes that at least some early systems are fused,
i.e. that separateness does not exist from the very beginning), since
neither of the two codes is yet fully or firmly established; the most
noticeable feature in the earliest stages is the abundance of instances of
mixing. The examples of different types of interference mentioned
below are taken from the speech of older children and adults.
5.1.1 Interference at the phonological level

Interference at the phonological level is often called ‘a foreign accent’.
It is usually more readily noticed by the layman than any other form of
interference. Adult bilinguals are more likely than child bilinguals to
show features of stress, rhythm, intonation and speech sounds from
their first language impinging on their second. It is relatively easy to
predict, in general, what kinds of phonological interference are the
most probable to occur in given groups of speakers, as the transferred
INTERFERENCE 97
elements are likely to be those that are absent in the other language or
dissimilar in the two codes (Lado 1957). For example, speakers of
Greek or Italian may tend to use vowel sounds similar to those in their
own languages when speaking German or English: Greek and Italian
have simple vowel systems, consisting of five and seven phonemes
respectively, whereas German and English have complex systems with
between sixteen and twenty units (including both pure vowels and
diphthongs). Thus native speakers of Greek and Italian may fail to dis¬
tinguish between long and short vowels, as in beat versus bit or Mus as
distinct from mu^. Similarly, a native speaker of English may carry
over the ‘dark’ (or velarized) [1] after vowels, as in halt, to the same
position when using German, which has only a non-velarised or ‘clear’
[1], thus realizing behalten as [bahakp] instead of [bahaltp]; and the
German speaker of English may fail to suppress his native final de-
voicing of plosives and produce only one form, e.g. [kit] for the two
English forms kid and kit, or [kAp] for cup and cub.
Interference in intonation can be a permanent feature in the adult
bilingual, as in the case of a Norwegian speaker who uses his native
rising intonation pattern at the end of sentences when speaking Eng¬
lish, which has the effect of making every statement sound as if he
were asking a question. In children, interference in intonation can often
be observed to adapt quickly when they change from one linguistic
environment to another.
5.1.2 Interference at the grammatical level

This kind of interference frequently involves such aspects of syntax as
word order, the use of pronouns and determiners, prepositions, tense,
aspect and mood. For instance, when Pascual, aged 4:2, asked his
father: ‘^Son esas las gafas que vas a verlo con?’ (instead of the well-
formed ‘^Son esas las gafas con que vas a verlo?’), he was using the
English word order with the preposition at the end ( Are those the
glasses you are going to see it with? ). An example taken from Cristi¬
na’s speech when she was 11:9 shows, again, the influence of English
word order on a second language, in this case German: Ich kann das
nicht kaufen, weil ich hab’ kein Geld mehr’ (‘I cannot buy it because I
have no more money’) (instead of ‘. . . weil ich kein Geld mehr
habe’).
Saunders, too, remarks on the fact that transference of word-order
patterns in his children’s speech was predominantly from English into
German, although he also gives some examples of occasional influence

in the reverse direction (1982a: 178).
One frequently reported type of interference concerns the word
order noun-adjective. Bilinguals who have Spanish or French as one of
their languages often use this order (noun followed by adjective), com¬
mon in Romance languages, when speaking non-Romance ones like
English or German. Among English—French bilinguals the reverse has
also been observed, that is, a preference for using the adjective-noun
sequence in French, which is possible but less idiomatic. (For a more
comprehensive treatment of the topic of acquisition and transference of
word order, see Romaine 1989: 187 ff.)
The following examples involving the use of the wrong preposition
still occur frequently in my children’s speech (at ages thirteen and ten):
Cristina: Tch fahre aufdem Bus nach Hause’

(instead of the correct ‘. . . mit dem Bus. . .’,
T go/am going home on the bus’)
Pascual: ‘Ich hab’ das auf dem Femsehen gesehen

(instead of the correct‘. . . im (= in dem). . .’
(T have seen it on TV’)
In both cases the phenomenon could be interpreted as interference

from English - after all, prepositions appear to be particularly suscep¬
tible ^to interference, as both observers of bilinguals and teachers of
foreign languages can testify. However, Suzanne Romaine’s discussion
of interference and prepositions (1989: 196 ff.) refers to studies of both
monolingual and bilingual children acquiring German, and some of the
examples given show either the overgeneralization of ‘in’ (in German
and English) or trouble in distinguishing English ‘in’ from ‘to’ and
German in from nach or an’. As monolingual children have prob¬
lems with prepositions as well, faulty production in bilingual children
cannot be explained wholly in terms of interference. According to Ro¬
maine, two issues have to be considered: the fact that prepositions are
often used with a variety of functions in different languages; and the
question of how this varied input affects (in the case of both monolin-
guals and bilinguals) the order of acquisition (Romaine 1989: 200).
The so-called ‘interference’, then, may well be due (at least partly) to
developmental factors.
INTERFERENCE 99
5,1.3 Interference at the lexical level

Interference at the lexical level is probably the most problematic ca¬
tegory for the linguist to account for. In the other types of interference
there can be little doubt that the criterion of ‘involuntary use’ could be
applied consistently and convincingly. But at the word level a different
situation seems to arise: the bilingual may borrow a word from one
language, when speaking in the other to a monolingual, either on an
ad-hoc basis or in a more permanent sort of way; and he may do the
same when addressing another bilingual. The borrowed item may even
be ‘fitted up’ with the appropriate morphology, as if to make the bor¬
rowing less obvious. For example, when she was seven years old
Cristina asked her father: ‘^Te vas a poner. . . te vas a poner el
beltoT (‘Are you going to wear. . . are you going to wear the belt?’
But ‘belt’ in Spanish is ‘cinturon’, i.e. nothing like belto). She had
used the proper Spanish word many times before and she quickly real¬
ized her mistake; this was a one-off occurrence of involuntary use. On
the other hand, the German for ‘zip’, which is more complex, ‘ReiB-
verschluB’, and ‘to zip up’, ‘den ReiBverschluB zumachen’, were less
firmly established in her vocabulary, and even at age twelve she still
used her own invented forms ‘Sippen’ (for ‘zips’) and ‘zusippen’ (for
‘to zip’) quite often. A fairly large number of ad-hoc formations which
clearly showed the influence of the other languages emerged when, at
various times, our children were given vocabulary recall tests in Span¬
ish, German and English. They produced such items as, for instance,
German ‘die Axe’ (based on English ‘axe’) instead of ‘die Axt’, or
‘das Kalf’ (based on ‘calf, the animal) for the correct ‘das Kalb’, and
Spanish ‘las grapas’ (based on English ‘grapes’) instead of ‘las uvas’,
or ‘esos ticos’ (on the basis of ‘those sticks’) for ‘esos palos’. The
incidence of lexical interference increased considerably when the
children were aware of being tested, when a time limit was set and
when testing in the three languages was done in quick succession.
Another kind of word interference is the overextension of the
meaning of a word into the realm of the other language. This may
happen, in particular, in the case of idiomatic expressions and items
with cognate forms:
Cristina (7:2): ‘Tu’ mir mehr Creme auf die Erdbeeren!’

(meaning ‘Give me some more cream for my strawber¬
ries’; but whereas ‘cream’ in English can be of either
the edible or the beautifying kind, German ‘Creme’
tends to denote the latter only - unless it occurs in a

compound like ‘Vanillecreme’.)
Cristina (7:3): ‘Und da haben sie sich den Kopf abgelacht’ (meaning
‘And then they laughed their heads off’, for the correct
German ‘Und da haben sie sich kaputtgelacht’)
Another example was provided by a German boy wishing a student

good luck with his exams, who said: ‘I’ll press my thumbs for you’
(meaning ‘I’ll keep my fingers crossed for you’, which in German is
‘Ich driick’ dir die Daumen’).
In these instances the interference was clearly from the bilingual’s
stronger language into the weaker one. In the case of older bilingual
speakers, e.g. immigrants who have lived for years in their adopted
country, the second language may start to influence the mother tongue,
so that native speakers may come up with cases like the following,
taken from the normal speech of a Spanish lady who has lived in Eng¬
land for most of her adult life: ‘jQue bien miras hoy!’ which is quite
impossible in Spanish, being the loan translation of English ‘How
good/well you look today!’ (‘mirar’ can only be ‘to look at’).
5.1.4 Interference in spelling

Interference in spelling is the transfer of writing conventions from one
langijage to the other. The following example shows that the bilingual
child who wrote this note had, at the time (when aged 11:2), imperfect
knowledge of German orthography and that, for the most part, she was
following the rules of spelling in English (which she knew better) and
her perception of spoken German: ‘Canst du bitte die watte
weknehmen? Die hamster machen viel crach wen sie im rat laufen. Ein
hamster ging grade em, aber 2? Ich hoffe daB du viel spaB gehapt
hast.’ (In the correct spelling this would be: ‘Kannst Du bitte die Watte
wegnehmen? Die Hamster machen viel Krach wenn sie im Rad laufen.
Ein Hamster ging gerade eben, aber 2? Ich hoffe, daB Du viel SpaB
gehabt hast.’ Two spelling conventions were therefore ignored: nouns
spelt with initial capitals, and in letters the pronoun ‘du’ is usually
‘Du’. Also, because the pronunciation is followed closely, words which
are subject to final devoicing or assimilation are misspelt, e.g. ‘weg¬
nehmen’, ‘Rad’, ‘gehabt’, ‘gerade’, ‘eben‘; and the sound [k] is
represented as ‘c’ instead of ‘k’. (The note explains that when the
mother had gone out and the girl went to bed she stuffed some cotton
BORROWING 101
wool into her ears so as not to hear her hamsters, and she was asking
her mother to remove it.)
5.2 Borrowing
All languages borrow lexical items from other codes, and have always
done so. In the European context it can be said that certain languages
seem to have been particularly prone to borrowing from others, as for
instance German, which has over the centuries incorporated large num¬
bers of words from Latin, Italian, French and, more recently, English.
English, too, has over the centuries borrowed extensively from other
European languages; today it is the most prolific ‘donor’, giving words
to most languages in Europe and beyond, often replacing indigenous
items which, from a linguistic point of view, were perfectly acceptable.
Just a few examples:
German der Computer [k3mpju:t'e]

Aids re:ts]
gestylt [ga'Itallt] ‘styled’
French le weekend [wik'end]

le parking [paRkig] ‘car park’
une star [staR] ‘famous actor/actress’
Spanish un mitin [mitin] ‘meeting’

el estres [estres] ‘stress’
los lideres [lideres] ‘leaders’
Borrowed items can be adapted either phonetically only (as with

French ‘star’ or Spanish ‘estres’) or both phonetically and morphologi¬
cally (cf. the German formation ‘gestylt’, the plural ‘lideres’) - or an
effort may be made to copy the native pronunciation (i.e. in the lan¬
guage of origin). It is interesting that ‘computer’ is pronounced in
German more or less in the same way as in English, even though the
sequence [pj] does not occur in any German word (one might have
expected to hear [komput-e] or something similar). There is often, in
addition, some change in the meaning in the borrowed item, even if it
is only in the context in which it is used: thus ‘gestylt’ applies to any¬
thing related to fashion and design, ‘un(e) star’ is often the leading role
in a film, ‘estres’ is seen in contexts about the demands or dangers of

modern urban life styles, ‘un mitin’ is mostly of a political kind.
Whether a borrowing (also called a ‘loan word’) becomes adapted or
not depends on a number of factors such as frequency of use, how
quickly it enters the general lexicon displacing a native word at least
partially, and how easy it is to integrate it phonologically and gram¬
matically. The process of adaptation and integration normally takes
some time to be completed. The fact that borrowings are very frequent
is reflected in the number of lists, sometimes in the shape of diction¬
aries, that continue to be compiled in many languages, as well as
numerous articles on the issue.
Loan words can be assigned to the ‘langue’ as described my Mac-
key (1970). Grosjean (1982) uses the term ‘language borrowing’ to
refer to terms that have passed from one language to another and have
come to be used even by monolinguals, and he distinguishes them
from instances where the bilingual borrows items spontaneously and
adapts their morphology, which he calls ‘speech borrowing’. The latter
clearly fall into the realm of ‘parole’, and they are unlikely to become
a permanent feature of the ‘host language’ (i.e. the language into which
they are borrowed). Grosjean gives (1982: 309) a delightful example
uttered by a French-English bilingual: ga a poppe (‘it popped’). Some
of the cases mentioned earlier under ‘lexical interference’ could be
treateji as speech borrowings. Many examples are included in Fantini’s
list of lexical borrowings (1985: 147 ff.) used by his son in both Eng¬
lish and Spanish, as for instance:
Mario (3:6): ‘Un juguete para el baby'
(‘a toy for baby’)
Mario (6:3): ‘Yo lo voy a lokar'
(‘lam going to lock it’)
(the Spanish is ‘cerrar con Have’)
Fantini proposes a subdivision into ‘pure borrowings’ (‘baby’, prob¬
ably pronounced as in English) and ‘adjusted borrowings’ (‘lokar’,
from English ‘lock’ with the Spanish infinitive ending coinciding with
that of ‘cerrar’).
Speech borrowing occurs in adults as well as children. The bor¬
rowed item serves a momentary need that may be caused by laziness,
fatigue or some form of emotional stress which makes the bilingual
forget the correct term. But the reasons for borrowing are not always
of a negative kind. A speaker may consciously choose an item from the
MIXING 103
Other language because (s)he considers it more appropriate or more to

the point. Incidences of borrowing in the speech of bilinguals directed
at other bilinguals do not normally lead to misunderstanding - on the
contrary, they can add interest, humour or intimacy to the conversation
and cause delight to both interlocutors at their shared linguistic knowl¬
edge. At the family level there may be a number of items borrowed
from one language into another; these tend to be related closely to a
particular culture and to have no equivalents: German ‘Abendbrot’ (‘a
cold evening meal’), Spanish ‘tapas’ (what one eats while one drinks
an aperitif) and English ‘Wakey, wakey!’ are among those used in our
household.
5.3 Individual creations
In the same way as monolingual speakers do, bilingual children can

come up with new, idiosyncratic linguistic creations and novel forma¬
tions. They may be the source of either confusion or laughter, and
sometimes they are adopted by the family as in-words. Bilingual child¬
ren can of course draw on the resources of two languages and use their
knowledge to produce forms such as, for example, English ‘knackber-
ries’ (to mean ‘snowberries’; the actual German word is ‘Knallerbsen’)
or German ‘hinterwarts’ (to mean ‘riickwarts’, analogous to English
‘backwards’). They can make up new forms in either language and, in
addition, use elements from both and combine them creatively. These
new terms are not the result of either interference or borrowing.
5.4 Mixing
‘Language mixing’ (or ‘code-mixing’) has been the object of consider¬

able attention in recent years. In approaching the study of this aspect of
bilingual competence linguists have usually taken one of two possible
broad lines:
(1) they have described the linguistic phenomenon as observed in

the speech of adults, focusing in particular on its relationship
with code-switching;
(2) alternatively, the discussion of mixing has centred on children’s
bilingual language acquisition and been studied in relation to the
question of language differentiation.
The large numbers of Hispanic Americans and English-speaking

Mexicans living in the USA has provided linguists with excellent op¬
portunities to study the language of speakers whose English contains
Spanish items and whose Spanish shows influences of English (e.g.
Gumperz and Hemandez-Chavez 1971; Hemandez-Chavez et al. 1975;
Timm 1975; Valdes Fallis 1976; Elias-Olivares et al. 1983; Poplack
1980; Silva-Corvalan 1982, 1983, 1986 and 1989). Mixing has been
studied in other parts of the world as well (see, for instance. Gibbons’
1986 comprehensive treatment of code-mixing and code-choice in
Hong Kong).
A good example (involving English and Spanish) of one code-mix
and two code-switches is given by Valdes Fallis (1982; 220): ‘And all
of a sudden, I started acting real curiosa (strange), you know. I started
going like this. Y luego decia (And then I said), look at the smoke
coming out of my fingers, like that. And then me dijo (he said to me),
stop acting silly.’
The following two are taken from another study of Chicano Spanish
(i.e. the Spanish spoken by some Mexicans living in the United States):
‘Pongan los picket signs’ (Set up the picket signs)

‘A veces, we take too many things for granted’ (sometimes,. . .)
(Reyes 1982)
Mixes can be used to clarify the message or make it more precise, a
point made by Silva-Corvalan that is illustrated in her example: ‘Tenia
zapatos blancos, un poco, they were off white, you know?' (he/she had
white shoes, a little. . .) (1989; 181).
Switches occurring at the lexical level within a sentence (intra-sen-
tential switches) are referred to as ‘code-mixes’ and ‘code-mixing’. On
the other hand, changes over phrases or sentences (inter-sentential), in¬
cluding tags and exclamations at either end of the sentence, are called
‘code-switches’ and ‘code-switching’ (Poplack 1980; McLaughlin
1984; Appel and Muysken 1987).
It may not always be easy, in fact, to see the difference between
speech borrowing and code-mixing in adults. At times there appears to
be a good deal of common ground between the two categories, the
only differences being of frequency and connotation, as mixing seems
to imply a more habitual feature and, possibly, one that is considered
somewhat negatively. The study of mixing in bilingual children, on the
other hand, touches on theoretical issues. The older bilingual child and
the adult bilingual are credited with possessing two different language
MIXING 105
systems, which they can keep apart; they can switch from one to the
other, and they can show code-mixes. But what about the two- or
three-year-old infant? Is he displaying one underlying compound sys¬
tem, or merely drawing on the resources of his two incomplete codes
when he utters examples like the following?
Thomas (2:2); (English-German bilingual, talking to his English

grandfather)
‘Lots of Moven, Grandad!’
(Saunders 1982a: 46)
Chr. (3:0): (German-French bilingual, addressing a French-speak¬

ing interviewer)
‘Non, das ist ein cadeau’
‘Non, da un chevaT
(Schlyter 1988)
And what about the even younger child, whose speech shows mix¬
ing at the one-word stage? As we saw in the previous chapter, linguists
engaged in research on young children’s mixing and language separ¬
ation have adopted two different theoretical positions, but for neither is
there conclusive evidence.
The working definitions used by researchers are often quite straight¬
forward. In the context of her study Schlyter proposes: ‘Mixing is
defined here quite simply as the child’s using words or sentences in the
‘wrong’ language, in a clearly monolingual situation; language separ¬
ation is defined as the opposite of mixing’. (Schlyter 1988: 2).
Redlinger and Park (1980: 339) write: ‘In this study, language mixing
refers to the combining of elements from two languages in a single
utterance’. Talking about definitions, Genesee (1989) suggests that: ‘It
is desirable to extend the definition of mixing to include single word
utterances from two languages during the same stretch of conversation
between a child and caregiver.’
The kind of mixes reported on may involve the insertion of a single
element, or of a partial or entire phrase, from one language into an
utterance in another, and they can be of a phonological (in the shape of
loan blends), morphological, syntactic, lexico-semantic, phrasal or
pragmatic kind.
Some instances of morphological mixing are:
papa’5 zapatos (= los zapatos de papa/Papas Schuhe)

(‘Daddy’s shoes’): German genitive form used in Spanish
die Ohren^ (= die Ohren/ears): English or Spanish plural form
used in German
The following contain lexical or phrasal mixes:

Tee schon gepourt (= Tee [ist] schon eingegossen) ‘Tea has
been poured’: English lexical item adapted into German morpho¬
logy
play mit water (= play with water): use of German preposition
male con Idpiz (= malen mit dem Bleistift) (‘draw with the pen¬
cil’): use of Spanish noun phrase with German verb
(examples from my children, at ages between 2:6 and 3:8)
And the froggie’s getting nass (‘wet’)
Das ist ein Knochen pour chien (‘that is a bone for dog’)
Le cheval ist zu miide (‘the horse is too tired’)
(examples from Redlinger and Park 1980: 340)
The following instance is from an English-Arabic child (aged 6:2).

It contains an English lexical part which has been morphologically
adapted to Arabic (example provided by my colleague Stephen Tho¬
mas):
X huwa be//v^ fii ilghaabi
he ‘lives’ in the jungle
The four children in Redlinger and Park’s study, and the three in
Suzanne Schlyter’s, were all in their second year, and their language
development ranged from one-word utterances (Brown’s Stage I) to the
more advanced phase of grammatical complexity involving subordinate
clauses (Stage V). The elements most often replaced by the children
were nouns and frequently used expressions (e.g. guckmaU, ‘look!’),
followed by verbs, which were usually conjugated correctly and in
some cases adjusted morphologically. This pattern confirms other data
about bilingual language acquisition (e.g. Taeschner 1983; Eantini
1985; Hoffmann 1985). With regard to function words, there seems to
be a wide range of possibilities: articles, prepositions, conjunctions and
adverbs were all observed to have been mixed in various ways.
In this connection Vihman (1985) makes the interesting point that
the proportion of mixed functors (or function/grammatical words) in¬
creases drastically if one puts all types of such items into one category
MIXING 107
and then compares them with the frequency of content (or lexical)
words. In the study of her bilingual Estonian-English son, Marilyn
Vihman found that in the early stages of language production he relied
heavily on English functors (25 per cent of the time) when speaking
Estonian, although 61 per cent of these were matched by an Estonian
equivalent which he had already learnt or was on the brink of acquir¬
ing. The author gives phonological and morphological reasons for the
occurrences, pointing to the greater complexity of the Estonian items
in comparison the English equivalents.
Mixing may occur for a number of reasons, the most important of
which can be summarized as follows.
(1) If an item has been acquired in one language but not yet in the
other, the child may use the one device (s)he has available to
express a certain lexical or grammatical meaning.
(2) If an item is temporarily unavailable, the subject is likely to re¬
sort to an equivalent form in the other language (or what (s)he
thinks is one).
(3) If an item is more complex, or less salient, in one language, the
young bilingual may make use of the corresponding one from the
other.
(4) If the child is exposed to mixed input (s)he will often respond
with mixed production.
With respect to this last point, Genesee found that many of the
studies he re-examined contained scant reference to adult input. More
often than not, researchers would accept their own general impressions,
or reports from parents, that the languages were used separately, when
in fact they were not. And he adds: ‘Evidence that mixing by bilingual
children can be traced in part to mixed input would weaken arguments
that mixing during early bilingual development necessarily reflects an
underlying undifferentiated language system’ (Genesee 1989).
Observers of bilingual language development agree in their reports
that mixing diminishes as the child gets older, but they differ in the
explanations that they offer, again reflecting the divergent theoretical
stances that they have adopted. By some of them, a decrease in mixing
is seen as evidence of the gradual separation of the child s initially
mixed language system. But one can also argue that, as the child s
languages develop, her vocabularies become more extensive and her
other linguistic resources reach a higher degree of sophistication, so
there is less need for her to borrow. A drop in the frequency of mixing
can also be explained in terms of the young bilingual’s growing aware¬

ness of sociolinguistic norms and her greater susceptibility to linguistic
clues in her environment.
Some studies have tried to find out why some children mix more
than others and why there is a difference in the pattern of their lan¬
guage separation. A number of inter-connected factors are seen to be
operating, which can be classified as relating to;
(1) the language input

(2) the linguistic development
(3) the general cognitive development.
(1) A number of linguists stress the importance of consistent and well-

balanced (both quantitatively and qualitatively) language input. There
seems to be general agreement that children raised on the one-parent-
one-language rule mix less than those who acquire their languages in
fused contexts. Arnberg and Amberg (1984) remind us that it is also
important to expose young bilinguals to monolingual speakers of both
languages. The monolingual interlocutor, they argue, is less likely to
understand a child’s mixed language output, thus forcing him to look
for the ‘correct’ item or try to paraphrase what he wants to say.
Schlyter’s (1988) study shows that mixing, in the German—French
children she considered, was affected by such factors as visits to
Frant;e or from French-speaking relatives. Increased exposure to the
language resulted in less mixing of German elements in their French.
(2) The linguistic factors influencing differentiation are those relating
to certain aspects of language development. The longer and more com¬
plex the child’s utterances become, the better he will be able to
distinguish his two codes, as he becomes more familiar with them.
Schlyter suggests in her study that for the children she observed this
language separation took place somewhere along Stages II or III, i.e.
the children showed a sharp decrease in their rate of mixing as they
progressed from the two- to the three-word stage. ^
(3) Vihman (1985) discusses the importance of cognitive development
for language separation at some length, particularly the interplay of the
child s increasing general self-awareness and his growing perception of
the two language systems. While not denying or playing down the sig¬
nificance of a child’s overall intellectual development for all aspects of
his maturation process, including his linguistic progress, it must be ad¬
mitted that the relationship between cognitive development and
CODE-SWITCHING 109
linguistic progression is a problematic area to assess, since the former

is measured, at least partly, by linguistic means (for example, by look¬
ing at how and how often the child refers to himself)-
We can say, then, that all children go through a stage of mixing.
The linguistic items that they mix may be known to them but tempo¬
rarily not available, or they may genuinely be known to the child in
only one of his languages. Input and situational factors (such as one¬
sided exposure or distractions) can influence mixing in terms of both
quality and quantity.
There are, however, many well-intentioned but uninformed people
who often become alarmed when they hear a child’s first mixed-lan¬
guage output, because they take it as a sign of the child’s inability to
cope with two languages and as a warning that there is some risk that
(s)he may get confused or retarded in her/his linguistic or cognitive
development. As a result, parents are sometimes advised against en¬
couraging their children to use more than one language during infancy.
The consequences can be that the child will then have to figure out
why the parents and family have suddenly changed their linguistic be¬
haviour. They may also be deprived of the possibility of benefiting
from the intimacy that a shared language provides, and from the wider
social and cultural contacts with other speakers of the parents’ lan-
guage(s). Bilingual children are always likely to be viewed differently
from monolinguals by society at large (if they live in a mainly mono¬
lingual society), because of their parents’ different background. If this
disadvantage is going to be there anyway, they might as well be
allowed to get the full benefit of being familiar with two cultures and
languages.
5.5 Code-switching
It might be argued that code-switching is potentially the most creative

aspect of bilingual speech. It has, however, also been considered as a
sign of linguistic decay, i.e. evidence that bilinguals are not capable of
acquiring two languages properly or keeping them apart. The topic has
certainly aroused the curiosity of many researchers, and a host of dif¬
ferent aspects of it have been studied. Sociolinguists have looked into
spieech communities, both monolingual and bilingual, trying to estab¬
lish reasons for, and patterns of, changes of style and language
switching. The bilinguals who have received their attention often in¬
clude immigrant groups like Hispanic speakers in the USA (e.g.
Gumperz and Hemandez-Chavez 1972 and 1975; Valdes Fallis 1976;

Silva-Corvalan 1983), Estonians in Sweden and the United States
(Oksaar 1974) and both minority and majority speakers in minority
areas, such as Catalan and Spanish speakers in Catalonia (e.g.
Calsamiglia and Tuson 1984) or Alsatian speakers in Alsace (Gardner-
Chloros 1985). Linguists have been more concerned with the question
of the linguistic status to be assigned to code-switching, whether one
can establish different types of switches and what kinds of constraints
can be at play, if any (see, for example, Pfaff 1979; Poplack 1980;
Aguirre 1985; and, for a good summary of the issues involved, Appel
and Muysken 1987).
5.5.1 What is code-switching?

The most general description of code-switching is that it involves the
alternate use of two languages or linguistic varieties within the same
utterance or during the same conversation. In the case of bilinguals
speaking to each other, switching can consist of changing languages; in
that of monolinguals, shifts of style. McLaughlin (1984) emphasizes
the distinction between mixing and switching by referring to code¬
switches as language changes occurring across phrase or sentence
boundaries, whereas code-mixes take place within sentences and
usually involve single lexical items. Much of the data presented on
codfe-switching involve cases of single-word switches/mixes, (e.g.
Oksaar 1974; Mkilifi 1978; Grosjean 1982; Aguirre 1985). In Oksaar’s
examples switches at the word level were always morphologically
adapted, as in the following:
(From an Estonian-Swedish bilingual):

‘Moddunud aastal ma sain skdtti [skat’t’i] tagasi’
(‘Last year I got some taxes back’)
(‘skatt’ is Swedish for ‘tax’)
(From an Estonian-American):
‘Naa, Sven, Mtshaiker tahab liftV
(‘Look, Sven, a hitch-hiker wants a lift’)
(here the first switch is phonologically adapted, the second also
morphologically)
(Oksaar 1974: 494-8)
Grosjean gives the following example to illustrate the difference he
sees between code-switching (a) and borrowing (b):
CODE-SWITCHING 111
(a) ‘9a m’ etonnerait qu’on ait code-switched autant que 9a’

(b) ‘9a m’ etonnerait qu’on ait code-switche autant que 9a’
(Grosjean 1982: 308)
Both sentences mean the same (‘I can’t believe that we code-switched
as often as that’). Borrowing, for Grosjean, involves morphological
adoption, code-switching does not. This distinction expresses the
underlying belief that code-switching is part of the bilingual’s ‘parole’,
while borrowing belongs to his ‘langue’.
5.5.2 What types of code-switches are there?

The following examples show a variety of code-switch patterns:
(1) A Spanish-English bilingual:

‘I started going like this. Y luego decia (and then he said), look
at the smoke coming out of my fingers’
(Valdes Fallis 1982: 220)
(2) A French-English bilingual:

‘Va chercher Marc (go and fetch Marc) and bribe him avec un
chocolat chaud (with a hot chocolate) with cream on top'
(Grosjean 1982)
(3) An English-German-Spanish trilingual:

Mother: ‘Na, wie war’s beim FuPball?’ (‘How was the foot¬
ball?’)
Pascual (9:3): ‘Wir haben gewonnen. Unsere Seite war ganz
toll. Ich war der (‘We won. Our team was brilliant. I was . . .’)
goalie. I stopped eight goals. They were real hard ones. (And
turning towards the pan on the cooker he continued) ‘Was gibt’s
zu essen?’ (‘What are we eating today?’)
(4) A Catalan-Spanish bilingual, in reponse to a Spanish speaker

who had said ‘Y se van a molestar, ^no? No tienen por que’
(‘They are going to be annoyed, aren’t they? [But] there is no
reason why they should’):
‘Bueno! Si que tienen por que. (‘Well! They do have a reason’)
‘O sigui, o I’encenen amb nosaltres o . . .’ (‘I mean, either
they have the bonfire with us or. . . ’)
(Calsamiglia and Tuson 1984: 114)
112 FEATURES OF BILE^GUAL SPEECH
(5) A Spanish-Catalan bilingual:

‘Hay cuatro sillas rotas y’ (‘There are four broken chairs and’)
prou!' (‘that’s enough!’)
(Ibid.: 115)
Code-switching can also occur at the phonological level i.e. when

the speaker changes the pronunciation patterns, as in:
(6) Cristina (4:0): (introducing her younger brother to a new friend

of hers)
‘This is PascuaT [paskwal]
Friend: ‘What’s his name?’ (i.e. she didn’t catch it)
Cristina: ‘Pascuall’ [paskwaef]
Friend: ‘Oh. . . ’
(The switch here consisted of changes in the vowel sounds from [a] to
‘schwa’ or [a] and, in the consonants, from initial unaspirated to aspir¬
ated [p'] and from non-velarized to velarized [9, to follow a more
English pattern.)
(7) An adult Spanish-Catalan bilingual:

‘. . . y si dices “perdon” en Castellano, se te vuelve la mujer y
te dice:’ (‘. . . and if you say “sorry” in Castilian Spanish, the
lady turns to you and says:’)
^ ‘En catald' (‘In Catalan!’)
(Calsamiglia and Tuson 1984: 115)
(8) An adult Spanish-American English speaker:

‘. . . Oh! Ay! It was embarrassing! It was veiy nice, though, but
I was embarrassed!’
(Silva-Corvalan 1989: 185)
(9) An adult Spanish-English bilingual:

‘Tenia zapatos blancos, un poco, they were off-white, you know.’
(Silva-Corvalan 1989: 181)
These examples illustrate the fact that there are many types of code¬
switches. Examples (1) and (2) contain switches within a sentence
(‘intra-sentential switches’), while in (3), (7) and (9) the switch occurs
between sentences (‘inter-sentential’); number (3) also involves a word
within a sentence, ‘goalie’. In (4) Spanish is used so as to establish
continuity with the previous speaker (who spoke in Spanish); once this
was done, the Catalan speaker continued in Catalan. (5) and (8) illus-
CODE-SWITCHING 113
Irate what Poplack (1980) calls ‘emblematic switching’, items such as

tags Cprou') or exclamations (‘ja}'!’) serving as an emblem of the bi¬
lingual character. In (6) the switch involves a change of pronunciation
features.
5.5.3 Who code-switches?

Code-switching has been observed to occur in the speech of children
as well as adults, for instance by Cornejo (1973); Padilla and Liebman
(1975); and Lindholm and Padilla (1978 a + b), who point out that it
begins to happen after the bilingual has become aware of speaking
different languages, i.e. it is not seen as part of early language mixing.
McClure (1977) reports that children tend to use various kinds of code¬
switches, depending on their age: whereas the younger children she
observed employed, for the most part, just English nouns in their Span¬
ish, the older ones would switch over phrases and sentences, and
would also use mixes at the word level. This suggests that a certain
level of linguistic proficiency has to be reached before bilinguals are
able to switch in the middle of an utterance. Poplack (1980) confirms
that only the fully bilingual Puerto Ricans she studied were able to
code-switch.
Bilinguals, particularly older ones, are normally able to control the
amount of code-switching they do. Many contextual, situational and
personal factors influence the speaker. In an informal conversation be¬
tween people who are familiar with each other and have a shared
educational, ethnic and socio-economic background, code-switching
can occur quite frequently. On the other hand, in a formal speech situ¬
ation between persons who have little in common code-switches may
be avoided because factors relating to prestige, language loyalty and
formality influence the language behaviour in such a way as to concen¬
trate the mind of the speaker on trying to approximate or keep to
monolingual standards. Bilinguals also differ among themselves in
their attitude to code-switching, both their own and other people’s.
Some have a relaxed disposition towards it, others consider that to
code-switch is a linguistic impurity or a sign of laziness and therefore
try to avoid it or correct themselves when they realize they have code-
switched; the latter group are also likely to signal intolerance towards
their bilingual interlocutor’s code-switching.
5.5.4 Can code-switching occur anywhere?

The two basic types of code-switches are those found across sentence
boundaries (inter-sentential) and those occurring inside sentences
(intra-sentential), as we have seen. Some efforts have been made to
find out whether there are likely to be universal constraints, or lan¬
guage-specific ones, on where switching can occur, and several
possible lines of exploration have been singled out. But there are cer¬
tain methodological problems that make it inadvisable, at the moment,
to try to make valid general statements on these issues. One area of
difficulty refers to the fact that the pattern of language switching may
well be determined by the specific pair of languages involved. Another
stumbling block lies in the nature of the research techniques employed:
in order to find out where in sentences switches are acceptable, re¬
corded material is analysed and bilinguals are asked to pass judgement
about possible switches; but sometimes these opinions have been found
to be in contrast with the comparable findings of other studies.
Some constraints on code-switching that have been proposed
(mainly in studies involving Spanish-English bilinguals in the USA)
are:
(1) a switch can occur within a noun phrase but only after the deter¬
miner, e.g. ‘se lo di a mi grandfather', but not ‘se lo di a my
V Grandfather' (T gave it to my grandfather’);
(2) adverbial constructions may be switched, as in ‘Vamos next
week' (‘We’ll go next week’), but not as interrogatives; ‘When
vamos' does not appear to be acceptable;
(3) an adjective may be switched after an adverb: ‘es mny friendly',
but it is not acceptable to switch the adverb in front of the adjec¬
tive (‘Es very amistoso’).
(All three examples are from Aguirre 1985.)
Timm (1975) proposes that an auxiliary and a main verb, or a main

verb and an infinitive, must be in the same language. On the other
hand, switches are possible between the antecedent and the relative
clause: ‘. . . those friends from Mexico que tienen chamaquitos'
(‘who have little children’) and also between the subject and the predi¬
cate in a copular construction: ‘. . . and my uncle Sam el mas
agabachado' (‘is the most Americanized’). (Both examples are from
Gumperz and Hemandez-Chavez 1975.)
Poplack (1982) suggests that a switch such as: ‘Mi hermano esta
CODE-SWITCHING 115
watching the game' is possible, and also one that involves phonological
adaptation of the main verb, e.g. ‘Mi hermano esta huachando the
game.' But a switch is not permissible between dependent morphemes,
so that something like ‘Mi hermano esta watchando the game' could
not occur (examples from Silva-Corvalan 1989: 181).
There is, however, ample evidence that bilinguals do engage in this
last type of code-switching as well:
‘Taipeo las cartas’ (‘I type the letters’)

'Puchamos el carro’ (‘We pushed the car’)
‘Va a mistir el tren’ (‘He is going to miss the train’, where an

epenthetic ‘t’ appears).
(Reyes 1982; 154)
These occurrences can be explained in terms of borrowing. As has

already been stated (in section 5.2), there is no clear dividing line be¬
tween borrowing and switching.
A third area of methodological issues involved in the analysis of
code-mixing concerns the notion of ‘host’ (or ‘base ) language: the lan¬
guage into which the switches are made. Appel and Muysken (1987:
121) remind us, however, that there are different ways of thinking of a
base language. Psycholinguistically, it is the speaker s dominant lan¬
guage, i.e. the one which determines his overall speech behaviour;
sociolinguistically, the base language should be defined as that of the
speech situation, that is, the unmarked linguistic form into which the
switch is made. In the above examples of Chicano code-switching,
Spanish would be the base language if psycholinguistic considerations
are given priority, but English is the sociolinguistic base language.
From a purely grammatical point of view, the base language may be
the one determining particular restrictions to specific switches.
5.5.5 Why do bilinguals code-switch?

The reasons for code-switching are manifold. Many of them (as illus¬
trated in section 5.5.2) are of a contextual, situational and personal
kind. Talking about a particular topic may cause a switch, either be¬
cause of lack of facility in the relevant register or because certain items
trigger off various connotations which are linked to experiences in a
particular language. Example (2) (page 111) shows this last point,
whereas (3) has a switch to ‘goalie’ possibly because the German item
was not known - and this switch then caused the rest of the utterance
to be in English as well.
Switching typically occurs when the subject is quoting somebody
else, as in example (7) (page 112), or being emphatic about something,
e.g. in (4) and (5). This kind of switch often takes the form of an
interjection, as in (8), or a repetition used for clarification, as in (9); in
both these cases the switches underline, in addition, the speaker’s per¬
sonal involvement and desire to be well understood. Contributions by
Oksaar (1974), Poplack (1980) and Calsamiglia and Tuson (1984) sug¬
gest that code-switching is also used to express group identity, i.e.
belonging to a bilingual community, like Estonians in Sweden or Puer¬
to Ricans in the USA, and solidarity with such a group.
Code-switching seems to be quite frequent among teenagers in immi¬
grant communities. This has been shown in studies carried out, for
instance, among guest worker communities in Germany (see Auer and
Di Luzio 1984). Hewitt (1982) indicates that even members of ma¬
jority groups (in this case, young whites in London) may switch (into
Jamaican Creole) in order to be accepted by a particular group.
Oksaar (1974) gives linguistic reasons for the code-switching she
observed. When asked why they had code-switched certain items for
which there existed acceptable Estonian equivalents known to them,
her informants gave her to understand that the Swedish or English they
used carried certain desirable connotations and reflected shared experi¬
ences, which was not the case in the other language. McClure (1977)
points out that children’s code-switches are often used with the inten¬
tion of clarifying the speech content for the interlocutor (example (6)
above shows this, too); only at a later age, perhaps eight or nine, does
the child begin to code-switch for emphasis or in order to focus on a
particular topic.
Code-switching, then, constitutes a habitual and often necessary part
of social interaction among bilinguals. Whereas monolinguals have
only one linguistic code at their disposal, bilinguals can rely on a four¬
way choice (the two languages and various forms of mixed and
switched codes, since they are able to code-switch in both their lan¬
guages). We can therefore represent the choices involved in the
bilingual’s speech behaviour schematically as follows;
CODE-SWITCHING 117
Bilingual
speaking to
monolingual monolingual bilingual

in LA in LB in LA and LB
wil' choose will choose will choose
LA LB LA or LB
with or with or
without without
elements of elements of
borrowing/mixing borrowing/mixing
/switching /switching
from LB from LA
Notes
1. Stage II is defined as mainly two-word utterances, no auxiliaries
or modals and hardly any grammatical morphemes. Stage III
consists mostly of three-word utterances, S/V/0 or S/V/Adv
structures, auxiliaries, modals, verb flexions, prepositions and
personal pronouns (Brown 1973). Schlyter (1988) suggests that
language-specific word order begins to emerge during Stage III,
although the children she studied showed slightly different mix¬
ing patterns. It is also during this developmental stage that the
input factors mentioned under (1) above play an important role.
Chapter 6
Cognitive and educational aspects of

bilingualism
6.1 Contradictory research findings
Many studies have suggested (see Chapters 10 and 11) that bilingual¬
ism can be established successfully without detrimental effects to the
child’s linguistic or personal development. Similarly, a generally posi¬
tive impression emerges from most research into bilingualism and
immersion education, carried out for instance in Canada, in Wales and
more recently in Catalonia. In immersion programmes involving initial
instruction (usually of majority children) in the L2, high levels of
achievement after some years, at no cost to the LI, have been reported
(e.g. Lambert and Tucker 1972). There is even some evidence that LI
skills may become enhanced by the experience (see below).
However, correspondingly positive results were not observed in
many minority children who underwent schooling in the majority lan¬
guage. According to Cummins (1976; 1978a; 1978b; 1978c; 1978d;
1979a; 1979b; 1981; 1984b; 1984c) and Skutnabb-Kangas (1976;
1977; 1979; 1984b), who spearheaded research into this issue in North
America and Scandinavia, children who had to use different languages
at home and at school (either because they were members of a linguis¬
tic minority or because they belonged to an immigrant family) often
showed an inadequate command of both the LI and the L2, and they
performed poorly in academic work.
Similar observations have been reported (Pfaff 1980; 1981) from
West Germany. The education of the children of migrant workers is not
necessarily more successful, nor is bilingualism established, by making
what is called ‘Vorbereitungsklassen’ available. These classes offering
CONTRADICTORY RESEARCH EINDINGS 119
immersion teaching of German are preparatory ones, and they are at¬
tended before the children join mainstream education.
How can this apparent contradiction be accounted for? It was noted
that Lambert’s (1974) distinction between additive and subtractive bi¬
lingualism helped to explain why studies of certain groups of children
(the first three of Skutnabb-Kangas’s four types - see section 2.3) pro¬
duced mainly positive results, whereas those in which negative
conclusions were obtained involved children in settings of subtractive
bilingualism. It has now become generally acknowledged that minority
children require the sort of educational provision that takes note of
both their social and their linguistic situation. Bilingual education pro¬
grammes - whether of the immersion or submersion (= ‘sink or swim’)
types, mother tongue or vernacular instruction - must consider a var¬
iety of social and attitudinal factors such as community support for the
programmes, relative prestige of the LI and the L2, parental and
teacher expectations, and the relationship of minority and majority lan¬
guage teaching. Bilingual education programmes in Sweden and, more
recently, also in Germany have shown considerably improved results
when certain pedagogic conditions were met, such as (1) the integra¬
tion of mother-tongue teaching and immersion teaching in the child’s
normal curriculum, (2) the presence of a mother-tongue teacher during
the immersion class, and (3) the teaching of the immersion course by a
bilingual teacher. Cummins (1978d) argued that the outcome of bilin¬
gual education (of whatever kind) appeared more or less successful
depending on
the developmental interrelations between a bilingual child s proficiency in
his two languages, his cognitive development and his academic progress.
In other words, in addition to specifying the regularities between societal in¬
puts and academic outcomes, it is necessary to pursue the connecting links
in the causal chain.
(Cummins 1978d: 396)
With the aim of establishing a theoretical framework for investigat¬

ing the ways in which the bilingual child interacts with his educational
environment, a central role was assigned to the developmental interre¬
lations between the LI and L2, and to the level of competence in both
languages. James Cummins has made a considerable contribution in
this area, putting forward a number of theoretical models. Some of
these are discussed below.
120 COGNITIVE AND EDUCATIONAL ASPECTS OE BILINGUALISM
6.1.1 Bilingualism and cognitive functioning

The debate about the relationship between bilingualism on one hand
and intelligence and various cognitive processes on the other has a
long history. Sometimes prejudice and passion were allowed to enter
into the discussion, particularly in the more distant past. Also, some of
the early views (pre-1920s or thereabouts) presenting a negative pic¬
ture of bilingualism were based on flimsy evidence and can be
discarded as unsubstantiated claims. Present-day research on the sub¬
ject can be grouped into three overlapping periods that Baker (1988)
terms:
(1) the period of detrimental effects

(2) the period of neutral effects
(3) the period of additive effects.
The question most commonly asked was: what effect does bilingual¬
ism have on intelligence? The assumption was that the causal chain of
events works in this particular sequence. It would of course be perfect¬
ly justified to ask the question the other way around, but the fact is that
this has not been done. Intelligence was seen as central to a child’s
*^9£qitiyg development and school success. It was thus cnnsidprpH
sential for parents to know about the nature of this relationship when
they took the decision as to whether or not to raise their children bi-
lingually. For teachers and people involved in education the issue was
of eveh greater importance since, so long as the view that bilingualism
had negative effects on intelligence prevailed, there could be no sup¬
port for any kind of bilingual programmes. The justification of
establishing bilingualism must rest, at least in part, on the conviction
that the child will benefit cognitively from being bilingual.
Some aspects of explorations of bilingualism and intelligence are
controversial. De Avila (1987), for instance, mentions the confusion of
issues which has ensued in some American research projects into mi¬
nority children s poor academic performance. He points out that low
school achievement is often explained by saying that these children are
at an academic disadvantage, as shown in intellectual, verbal, motiva¬
tional and cognitive style-differences - all of which have been equated
with bilingualism: ‘Unfortunately, studies offered in support of this
contention have tended to confound both poverty and ethnolinguistic
group membership with linguistic proficiency.’ (De Avila 1987: 150.)
Baker (1988) suggests that most research into the relationship be¬
tween bilingualism and intelligence fails to acknowledge that the
CONTRADICTORY RESEARCH FINDINGS 121
nature of the latter is a controversial issue in psychology, sociology,

genetics and semantics. He calls our attention to the difficulty of ob¬
serving and specifying intelligence in a totally satisfactory way, its
variably defined structure and the unresolved question of the extent to
which it derives from hereditary influences. Furthermore, he raises the
problems inherent in IQ tests, pointing out that they tend to be used
not only as tests measuring intelligence, but as predictors of future aca¬
demic achievement, thus changing their intended function.
Consequently, they may be less fair to minority than to majority child¬
ren: ‘IQ tests have tended to reflect and reinforce mainstream culture,
the educational status quo and a curriculum where convergent thinking
skills are highly valued. . . Given a more multicultural, creative and
progressive curriculum, IQ tests may have a much lower predictive
power’^ (Baker 1988: 7).
6.1.2 Overview of research on bilingualism and cognitive

functioning
(1) Period of ‘detrimental effects’ Nineteenth- and early twentieth-cen¬
tury philosophers, educators and philologists have repeatedly expressed
the belief that bilingualism has an adverse effect on the cognitive de¬
velopment of the child (see also section 7.1). This conviction was not,
however, based on empirical research. More influential on educational
thinking (and also, probably, on political decision-making) were the
results of a number of IQ tests carried out in English-speaking coun¬
tries where bilingualism was seen to cause problems, e.g. in the USA,
Britain (Wales and Gaelic-speaking Scotland) and Ireland. Specifically,
the difficulties encountered related to the education of minority or im¬
migrant children who, quite consistently, achieved lower IQ scores
than did monoglot ones.
Nowadays these tests are considered invalid. Contemporary standard
testing takes into account a number of additional factors, such as socio¬
economic background, school conditions and emotional aspects, which
have a bearing on marginality and low status. In the former tests in
question these variables were not controlled, and in most cases the
language of the tests was the one common to both monolinguals and
bilinguals, which usually meant that the latter were tested in their
weaker language. The tests themselves were also criticized for contain¬
ing, as they did in some instances, verbal items which would run
counter to a valid assessment of the members of the bilingual group. A
122 COC.Ni riVE AND EDUCATIONAL ASPECTS OF BILINGUALISM
tiiilhei controversial point is the inclusion of questions which mitjht be

seen as biased towards a certain class-culture.
From the 192()s onwards research was undertaken in Wales, often
by people who, according to Baker’s (1988) review, were in favour of
the use ol Welsh as a second language. Nevertheless, they obtained
results which appeared to lend support to the view that bilingualism
can have a negative effect on intelligence. The studies by Saer (1922;
1923) were carried out, in rural areas and small urban districts in South
Wales, on a large number ot children and (another set of tests) on
college students. Both English and Welsh were used as test lamiuases.
Few ditteiences were found between the scores of urban bilinguals and
urban and rural monoglots, but rural bilinguals had lower scores, and
they were found to be two years behind peer monoglots. Saer con¬
cluded tiom his investigations that not only were rural bilinguals
disadvantaged I’/.v-d-v/.v English monolinguals, but in addition they
weie mentally confused, a negative consequence of their bilingualism
which was likely to stay with them for the rest of their lives. These
findings were taken very seriously by the school authorities. The>-
proved an obstacle for Welsh mother-tongue teaching and bilingual
education, but they also became the subject of intense scrutiny. The
upshot was that a number of methodological errors were detected, thus
questioning the validity of the conclusions (for a detailed discussion,
see B;i|;er 1988).
(2) Period of ‘neutral ejfeets' This stage overlaps both with the above-
mentioned one and with the period of ‘additive effects’. Its main
contribution was the various reviews of investigations on the subject
that It produced. For instance, the articles by Arsenian (1937) and
Darcy (1953) reported on certain inadequacies in the methixls pa'-
viously employed. But. whereas their general conclusion was that
bilingualism itself does not affect intelligence, they. tCK'». repcirted on
the imperfect knowledge of the L2 possessed by many bilinguals,
which was apparent in tests requiring the use of verbal skills.
Studies carried out in the 1950s by W. R. Jones in Wales (e.g. Jones
1953; 1955; 1959) were concerned with measuring the intelli4nce le¬
vels of a large number of bilingual childmn. At first they appe.ued to
show a correlation between monolingualism and higher intelligence
scores. However, after re-analysing the data and a-categorizin^' several
variables (tw high a pmportion of his bilinguals had a maniutl-gamp
background, so they had been over-represented, whereas most moniv
glots belonged to the non-manual group) he came to the conclusion

that there were no significant differences between the two as far as
non-verbal IQ tests were concerned. His final results, published in
1959, therefore point towards no effect of bilinguali.sm on intelligence.
Jones concluded that bilingualism was not necessarily a source of intel¬
lectual disadvantage, and that there were no significant differences in
IQ scores between bilinguals and monolinguals if the variable ‘parental
occupation’ was taken into account - which presumably means that the
differences between the two groups were attributable to factors arising
out of socio-economic dissimilarity.
(3) Period of ‘additive effects’ A turning point in the debate was

reached with Elizabeth Peal’s and Wallace Lambert’s investigations in
a Canadian setting (1962). Starting again.st the background of the
mainly negative nature of previous re.search, they expected to find a
bilingual deficit in their subjects. Their aim was ‘to pinjxiint what the
intellectual components of that deficit might be in order to develop
compensatory strategies’ (Lambert 1977: 16).
Peal and Lambert were aware of the shortcomings of previous re¬
search, and they took care to control the important variables. Their te.st
groups (all children came from middle-class French schools in Mon¬
treal) were matched for .socio-economic class, gender, age, school
grades and language proficiency, and only balanced bilinguals and
monogloLs were included, thus reducing their original .sample of 364 to
110. Their findings were surprising. In Baker’s words, they ‘ . . .
provided an appetizer, .stimulant and menu for future research’ (Baker
1988: 17).
They found that bilinguals scored more highly than monolinguals in
both verbal and non-verbal measurements of intelligence. The authors
argued that the former had a more diversified structure of intelligence
and greater mental flexibility, and that therefore the cognitive function¬
ing of bilinguals benefited from their bicultural experience and from
positive transfer between languages.
These conclusions have, in turn, been critically examined and found
to crmtain a number of weaknesses. The facts that the children tested
came from middle-cla.s.s homes and that only balanced bilinguals were
selected made it doubtful whether the subjects could be seen as repre¬
sentative of bilinguals in general. MacNab (1979), and others after
him, noted that choosing children according to parental occupation or
sfx:io-ecrjnomic background might well mean that factors such as cul-
124 COGNITIVE AND EDUCATIONAL ASPECTS OF BILINGUALISM
tural milieu and social environment were not adequately accounted for,
although they could be of cmcial significance for a child’s language
experience. Another problem area concerned researchers’ ways of ar¬
gumentation. If test results show that there is a correlation between
bilingualism and intelligence, this does not necessarily imply a causal
link in either one direction or the other. Baker (1988) suggests that
there may even be a third dimension to the ‘chicken and egg game’,
namely that one (IQ) is simultaneously the cause and effect of the
other (bilingualism); in his words, there is ‘ . . . mutual interaction
and stimulation between dual language and cognitive abilities’ (Baker
1988: 19).
However, Peal and Lambert’s research had a profound impact on
studies in this field. The techniques that they used were sophisticated
and their tests well constmcted. Their findings became widely known
and led other people to look for positive effects of bilingualism on
various other aspects of cognitive functioning. Some of these findings
are outlined below.
In the context of research on cognitive development, an observation
by Leopold (1949a), which is found in his case study of child bilin¬
gualism, appears to be relevant. He states that his daughter was not in
the habit of clinging to particular words in songs, rhymes and stories,
as is normal in young children; instead, she would readily accept new
names for objects, and she did not repeat her favourite stories with
stereotyped wording. Leopold attributed this looseness of the link be¬
tween words (i.e. their phonetic shape) and their meaning to her
bilingualism. A special sensitivity to language, manifested in the form
of remarks on different forms of conveying meaning (including the use
of other codes), was also noted by Imedadze (1967) and Fantini
(1985), while Vygotsky (1962) had already argued that the positive
side of bilingualism was that it led to greater awareness of linguistic
operations.
In the 1970s a number of studies tried to establish the extent to
which bilinguals were superior to monoglots in certain aspects of cog¬
nitive functioning, such as divergent and creative thinking and
analytical orientation towards language. Anita lanco-WorralTs (1972)
study was carried out in South Africa with English-Afrikaans bilin¬
guals whom she tested for metalinguistic ability. Only in the tests
where the children had to make choices according to semantic or
phonetic criteria did she find a significant difference between monolin-
guals and bilinguals. For example, in a test in which they were asked ‘I
have three words, CAP, CAN and HAT - which is more like CAP,
CAN or HAT?’, the bilinguals tended to select the semantically related
word, but virtually all the monolinguals chose according to sound.
Among the older children, both bilinguals and monolinguals tended
towards a choice according to meaning. It could therefore be con¬
cluded, she claimed, that the bilingual children were at a more
advanced stage of metalinguistic awareness, i.e. they had a greater
ability in their consciousness of language forms and properties. Cum¬
mins (1978a) extended lanco-WorralTs tests to Irish-English and
Ukranian-English bilinguals and found similar results.
Sandra Ben-Zeev’s experiments (1976) were carried out with He-
brew-English children (from Israel and the USA) and with
Spanish-English bilinguals in the United States. Her tests involved
word associations and word substitutions. One item, for instance, was
in the form of a ‘game’ in which it was decided to call an airplane a
‘turtle’, and the question to be answered was ‘Can the turtle fly?’ (cor¬
rect answer: yes). A more complex version entailed sentence
production with a substituted word that resulted in a violation of con¬
ventional rules of grammatical agreement, e.g. in the game the word
‘macaroni’ had to be used for the first person pronoun ‘I’, so that the
question ‘How do you say “I am warm”?’ was supposed to be
answered ‘Macaroni am warm’. In her various tests she found that bi¬
linguals obtained better scores in word substitution exercises and also
when they had to disregard grammatical rules of sentence constmction
(e.g. they avoided incorrect - in the game - replies such as ‘/ am
warm’ or ‘Macaroni is warm’). Ben-Zeev inferred that bilinguals
showed greater cognitive flexibility and were capable of more complex
analytical strategies in their approach to language operations.
Sheridan Scott’s (1973) study in Montreal aimed at finding a link
between bilingualism and divergent thinking. Her study differed from
others in that the children she considered were in the process of be¬
coming bilingual (i.e. they were not bilingual already, before the
testing began), and her approach was thus geared towards finding out
whether their becoming bilingual would have any impact on their cog¬
nitive development. She compared two groups of children: one was
given the opportunity of becoming bilingual in English and French
during the seven-year period of the testing, by means of an immersion
programme; the other group followed a conventional English-language
curriculum. The two groups had been matched for IQ levels and social
class. Again, her results point towards positive effects, because the bi-
JG
linguals were the higher scorers in tests which required the subjects to
focus their attention on something, e.g. a paper clip (‘think of a paper
clip ), and then asked them to list all the things that one could do with
it which came to their minds.
Both lanco-Worrall (1972) and Ben-Zeev (1976) suggest that bilin¬
guals may possess greater sensitivity towards verbal and non-verbal
feedback cues than monolinguals. This point has been taken up by
other researchers (e.g. Bain 1976), and it finds expression in some of
the ideas put forward by Cummins (1976). The argument is that bilin¬
guals have a wider and more varied range of experience than
monolinguals, as they have access to two cultures and operate in two
different systems. Even if this does not result in doubling the total
range of social and cultural experience, there may be a potential gain
for bilinguals. Their need to switch from one code to another has also
been seen as beneficial to flexible thinking, as each language may pro¬
vide the speaker with distinct perspectives. Although neither of these
ideas can yet be supported by conclusive research evidence, they do
suggest an interesting panorama.
A substantial amount of ‘positive effect’ research is going on at the
moment. However, as in the case of the less optimistic findings, close
inspection of aims and procedures can disclose shortcomings - often
concerning, once more, methodological issues. Many research schemes
are set in contexts of additive bilingualism where children receive en-
courag'fement about their use of two languages, and also about learning
in general. The problems of sampling and matching that seem to be
inherent m all testing involving bilinguals have already been men¬
tioned, as has the question of a possible causal relationship between
bilingualism and cognitive processes. Baker (1988) comes to the con¬
clusion that the trend of research appears to point towards positive
effects rather than negative ones, but he is prudent enough to add that
further advances in the field are needed before we can confidently
answer the questions of whether the extent of the advantages is signifi¬
cant and whether bilingualism actually fosters educational progress. We
shall have no real reason to be optimistic until such day as bilingual
children receive, as a matter of course, widespread encouragement and
some form of bilingual education, so that they can achieve their fiillest
possible potential in each of their languages.
LINGUISTIC COMPETENCE 127
6.2 Linguistic competence, cognitive functioning and the

education of minority children
6.2.1 Semilingualism - ‘BICS’ and ‘CALP’

The debate on semilingualism that started in Scandinavia in the early
1970s (see section 1.3) led to further discussion of what precisely con¬
stitutes competence in a given language. It was found that the children
of Finnish immigrants in Sweden frequently had poorly developed lan¬
guage skills in their home language when they started school, and that
their levels of proficiency in Swedish (then the only school language)
were also below those of comparable monolingual children in spite of
their average intellectual ability. Skutnabb-Kangas and Toukomaa
(1976) point to the social origin of the phenomenon: the children they
had observed tended to come from socially deprived families with an
umstimulating home-language environment, and they found themselves
compelled to acquire the L2. These researchers argue that .semilingual-
ism should not be seen as a purely linguistic concept, but as one that
includes cognitive a.spects as well. Thus, they use the term for children
who did not show ‘full command of Swedish’ (p. 28) in the measure¬
ments taken in vocabulary tests, taking also into account their
understanding of abstract concepts and synonyms (p. 21). The.se auth¬
ors remark on the discrepancy between the children’s apparent
linguistic competence (the impression of fluency they gave) and the
lack of conceptual linguistic knowledge that became clear only when
they had to perform in tests which required more complex cognitive
operations.
In his earlier work Cummins (1978c; 1979a; 1979bt had referred to
these two aspiects of linguistic competence as ‘surface competence’ and
‘cognitive linguistic competence’. By the former he meant ‘those
“visible” features of language which are relatively easy to measure,
e.g. pronunciation, vocabulary, grammar, fluency, etc. , whereas the
latter denoted ‘the ability to make effective use of the cognitive func¬
tions of the language, i.e. to use language effectively as an instrument
of thought and represent cognitive operations by means of language’.
When considering the relation.ship of the bilingual child’s LI and L2,
Cummins’ notion of language competence referred to the latter type.
In later publications Cummins distances him.self, just as others have
done, from the notion of semilingualism, becau.se of the ambiguous
and even pejorative associations the term came to acquire. But he
maintains his basic idea of language competence as consisting of two

different components. He distinguishes ‘BICS’ (‘basic interpersonal
communication skills’, which are those aspects of linguistic skills that
are necesSary for functioning in everyday contexts) from ‘CALP’
(‘cognitive academic language proficiency’, that is, the skills which are
required outside immediate everyday communication situations).
‘BICS’ are sufficient for cognitively less demanding situations, and
they can be measured by pronunciation and fluency tests. ‘CALP’
skills, on the other hand, are said to show high correlation with verbal
sections of intelligence tests (Skutnabb-Kangas 1984a). Cummins
(1981, 1984b) developed his ideas further by incorporating into lin¬
guistic competence the notions of context and cognitive complexity.
Thus, he sees some communicative tasks as being cognitively unde¬
manding, particularly when they are accompanied by paralinguistic
features and if situational clues can provide feedback. On the other
hand, certain activities (for instance, in the classroom) involve decon-
textualized language and are therefore of a more cognitively
demanding type.
Cummins argues that surface fluency, or ‘BICS’, can develop fairly
easily in minority children if they attend a classroom where a good
deal of contextual support for language learning is provided. Further¬
more, surface fluency in each language, LI and L2, can develop with
relative independence from the other. By comparison, context-reduced
and cbgnitively demanding communication develop interdependently.
This means that the kind of cognitive skills required for success in the
more intellectually demanding tasks must be dependent on the relation¬
ship between the LI and the L2, and that unhindered learning can take
place only if both languages are sufficiently well established. Cum¬
mins Developmental Interdependence Hypothesis suggests that the
level of second language competence a child acquires depends to some
extent on the stage of development that has been reached in her first.
This theory provides the rationale for the kind of bilingual education
that advocates the use of the child’s mother tongue during the early
stages of education, adding the second language only when she has
developed higher-order cognitive and linguistic skills in the first. The
assumption is that the ability to handle cognitively demanding tasks
can be transferred to the L2 successfully.
LINGUISTIC COMPETENCE 129
6.2.2 The Balance and the Think Tank models

Speculation about the nature of the relationship between the bilingual’s
two languages has produced many hypotheses. With respect to the
question of cognitive functioning and its educational implications, two
models have been particularly influential.
(1) The Balance theofy This model makes two basic assumptions. One
is that the brain has only a limited capacity and therefore the addition
of a second language automatically leads to a decrease of proficiency
in the LI. When commenting upon it, Cummins (1981) uses the ana¬
logy of balloons: the bilingual has two half-full spaces, depicted as
balloons, in his head, representing his LI and L2 language proficien¬
cies, while the monolingual has only one, better-filled balloon. The
other assumption concerns the storage of language ability: the LI and
the L2 are seen as separately stored systems, and it is considered that
they function independently of one another. Cummins refers to this as
‘separate underlying proficiency’ (or ‘SUP’).
This theory encourages the belief that bilingualism may result in
some form of linguistic deficit, and that cognitive and educational de¬
velopment may become impaired by the bilingual experience. It is
unable to account for code-switching and other features of bilingual
speech in any terms except as interference. In view of the substantial
amount of research which suggests probable positive effects of bilin¬
gualism on cognitive functioning, this theory has few advocates today,
although it may still be adhered to by some lay persons who view
bilingualism with suspicion.
(2) The Think Tank model Cummins (1981) made this proposal in re¬
sponse to the Balance theory. In place of the two balloons, he suggests
the metaphor of a Think Tank containing the bilingual’s LI and L2
proficiency. Each maintains its separate linguistic characteristics,
which are dependent on input and feedback received from output, yet
the same mental expertise underlies reception and production of both
languages. The notion of ‘common underlying proficiency’ (or ‘CUP’)
is crucial to this model, expressing as it does cognitive activity as
being centralized and independent from a particular language. Em¬
ploying both languages will foster the bilinguals’ think tank. However,
if forced to make use of the weaker one only, for instance in the class¬
room, the bilingual child may encounter problems in understanding
whatever is being taught, and this may in turn hold back the develop-
merit of the think tank and, consequently, lead to poor performance in

tasks involving skills in both the home and the school language.
This theory can explain the successful acquisition of bi- or multilin¬
gualism, and it has no problems accounting for phenomena of bilingual
speech. It is also congruent with other theories about the relationship
of language and educational attainment. Like the previous model, how¬
ever, it has certain limitations (see section 6.2.3 below).
6.2.3 The Threshold Theory

The Threshold Theory was developed in the attempt to deal with the
inconsistency in findings regarding bilingual children’s cognitive
ability. It was proposed by Skutnabb-Kangas and Toukomaa (1977)
and by Cummins (1976; 1978d), and based on their examination of
issues relating to the education of minority children. It was concerned,
especially, with establishing at what point bilingualism can be seen to
lead to positive cognitive consequences.
Two thresholds are proposed in the levels of the bilinguals’ linguis¬
tic competence. The lower threshold must be attained if negative
cognitive effects are to be avoided, as limited linguistic skills will
Type of bilingualism Cognitive effect
baianced bilinguals, positive

high levels in both cognitive
languages effects
higher thresh
"O
(U level of bilinc
> dominant bilinguals, neither
<D competence
O native-like level in one positive nor
of the languages negative effects
lower thresh(
300 level of bilinc
semilinguals, negative
c competence
ra low levels in both cognitive
JD languages (may be effects
dominant or balanced)
(D
Source: Toukomaa and Skutnabb-Kangas (1977: 29)

FIGURE 6.1 Diagrammatic representation of the Threshold Hypothesis showino
cognitive effects of different types of bilingualism
SUMMARY OF THE ISSUES 131
hinder academic progress and cognitive growth. If the child achieves a

level of bilingualism somewhere between the lower and the higher
threshold (i.e. if she can function effectively in at least one of her
languages), there are likely to be neither positive nor negative effects
on cognitive abilities. High attainment in both languages occurs be¬
yond the upper threshold. Such linguistic competence may entail
positive cognitive effects.
6.3 Summary of the issues
Cummins’ hypotheses concerning language proficiency and educa¬

tional success of bilinguals take into account the various types of
competence and skills required of pupils for performing different learn¬
ing tasks. Together with the Developmental Interdependence Theory,
the Threshold Theory is able to provide an explanation for the fact that
migrant children who become bilingual in subtractive contexts, and
who are taught through the medium of their weaker language, fail to
develop sufficient competence in their second language and are under¬
achievers in school tests. These models underline the requirement that
transitional education programmes (where the child is taught in her LI
before she joins mainstream education in the L2) must be sufficiently
long to permit the development of cognitive processes, which is
necessary for coping with school tasks. Equally, the theories may ex¬
plain why initial reception classes (for instance the ‘Aufnahmeklassen’
or reception classes in Germany, where Turkish children are taught
German before joining monolingual peer groups) may not provide the
hoped-for basis for migrant children to succeed in majority education.
Similarly, the models offer an explanation as to why immersion pro¬
grammes (such as the ones offered, on a voluntary basis, to majority
children in Canada) can be successful. Although there may be tempor¬
ary lags in achievement once the second language (French) is
developed to a level that enables the children to cope with concepts
through this medium, educational schemes that include immersion pro¬
grammes (in Canada) have not been reported as having had detrimental
effects on the children’s progress. Naturally, in other parts of the
world, where social and cultural conditions may be different (e.g. the
Netherlands, Germany, Wales, Catalonia or the Basque Country), im¬
mersion programmes offered to minority children may produce quite
different results.
Upper Level
Age-appropriate competence in two or more

languages
Second Threshold Level
Middle Level
Age-appropriate competence in one but not two

languages
First Threshold Level
Lower Level
Low level of competence in both languages:

Negative Cognitive Effects
First Language Second Language

Proficiency Proficiency
Source: Baker 1988, p. 176
FIGURE 6.2 Bilingualism, Cognitive Functioning and the Threshold Theory
For the linguist these theories are less satisfaetory than for the edu¬
cationalist. They provide a neat representation of bilingualism and
cognitive functioning, but although they use the notions of ‘levels of
bilingual attainment’ and ‘bilingual competence’ there are no concrete
examples of what these should be - for either ‘BIGS’, ‘CALP’ or the
thresholds levels. Baker’s (1988) representation of the Threshold The¬
ory provides us with a little more information than the original models.
He shows language competence separately for each language, and he
specifies the threshold levels by reference to an assumed ‘age-appro-
priate’ norm. So within each of the three levels one could represent an
individual’s bilingualism by indicating his or her language proficiency
in LI and L2 on the ‘rungs’.
In an article in which they review some of the research into semilin-
gualism with a view to reassessing this notion, Marilyn Martin-Jones
and Suzanne Romaine (1986) point out that in his definition of lan¬
guage competence Cummins appears to be equating semantic with
cognitive development. They argue that the relationship between lan¬
guage and thought processes may in fact be a great deal more complex
than Cummins seems to suggest, and that it is oversimplistic to claim
that the various language skills can be compartmentalized and tested
separately in the way that he proposes.
Another criticism relates to the vague and rather narrowly defined
concept of ‘school success’. The idea of what constitutes success ap¬
pears to be applied in the same way to all children, majority and
minority, yet it tends to be defined in terms of traditional, middle-class
values deduced from measured school tests, and thus reflecting ma¬
jority-group standards. The minority child’s perceived need and
motivation to leam the second language and her expectation of educa¬
tion are likely to affect the outcome of the whole range of school
activities.
Terms such as semilingualism and notions like ‘BICS’ and ‘CALP’
will remain vague as long as there are no reliable means of testing the
bilingual child’s whole bilingual competence. At the moment it is not
clear whether a supposed deficiency, as measured in formal tests and
shown, for instance, as simplified language using fewer morphological
devices, word-order irregularities or limited vocabulary, actually affects
communication. Some linguists (e.g. Stblting 1980; Grosjean 1985a;
Martin-Jones and Romaine 1986) have warned against proposing theo¬
ries that may be misused as deficit interpretations outside the original
remit within which they were conceived. It should be remembered that
the above-mentioned models take a narrow view of cognitive function¬
ing and bilingualism, as they do not allow for a wide range of cultural,
social and political factors that can also influence the school achieve¬
ment of minority children. Such aspects as the child’s motivation
towards learning in general, attitudes towards the L2, its speakers and
school, parental support and attitudes, and the expectations and dispo¬
sition of teachers regarding minority children may in fact be just as
influential in determining attainment as linguistic and cognitive factors.
They must be taken into account in interpreting the often observed
phenomenon that, within certain multicultural settings, children belong¬

ing to some minorities have a higher success rate in educational
achievement than those from other minority groups.
Lastly, we must remind ourselves that research findings present
average scorings, i.e. they level out individual scores which may show
considerable variation in relation to each other. Human beings experi¬
ence language, culture and education in various differing ways, and
whereas some may be more affected by cognitive or linguistic determi¬
nants, for others bilingualism and educational performance can be
influenced largely by factors of a motivational or attitudinal kind.
An attempt has been made in the last two chapters to show that one
cannot assume an a priori relationship between bilingualism on the one
hand and personality development and cognitive functioning or educa¬
tional achievement on the other. If the human mind can cope with, a
multitude of different functions and operations without suffering any
detriment, why should bilingualism represent an exception? Whatever
link between language and mind we want to postulate, it must be of
such an abstract nature that it is able to accommodate any number of
natural languages in this binomial relationship. There is no evidence
that bilinguals suffer any negative effects in their development on ac¬
count of their bilingualism, nor do they acquire a split view of the
world. The once-famous Whorf-Sapir Hypothesis, which claimed that
speaking a given language inevitably leads to holding certain cultural
values (in other words, that the language one speaks determines one’s
world view), has remained a hypothesis. It would certainly be difficult
to support it with regard to a bilingual - for which of his two lan¬
guages would determine his experience of the world? Or does the
bilingual have some kind of fluid values that can be adhered to or
abandoned according to the language used at particular times? This,
however, is not to say that bilinguals do not adjust their cultural beha¬
viour as they switch from one language to another, so as to comply
with the norms of communication and social conduct peculiar to each
speech community. But their view of the world remains unaffected by
this.
.Naturallyi.th?Jbdingual’s^^qgnitive__and__educational development is
^’X social and psychological factors, but so is thq
monolingual’s, even jf the process is manifested differently. Like all
children, bilinguals need parental support and favourable sjocmTattl^
tudes around them to maintain their motivation for learning. They have
special needs in so far as attention has to be paidjo both their Ian-
guages, at least until the two systems have become firmly established.
There are a number of bilingual education programmes that provide
support for the non-majority language, but only for a relatively short
period of time, which means that afterwards the maintenance of the L2
comes to depend on the home or community efforts. It is frequently the
case that the child remains bilingual but becomes dormnant in the lan-
guage of the rnajority and therefore establishes stronger (although not
necessarily exclusive) links with the culture of the latter. Older immi-
grants or socially stigmatized migrant groups are more likely to suffer
from psychological problems, brought about by home-sickness, feel¬
ings of guilt at abandoning previous cultural values, low self-esteem,
social isolation, unemployment, poor housing and health-care - the hst
is long. However, these are not problems resulting from bilingualism,
_but rather ê consequences of migration or minority status.
Notes
1. Traditional IQ tests require a correct reply for each question.
This focus on one single acceptable answer has been termed
‘convergent thinking’. ‘Divergent thinking’ represents a more
open-ended skill which allows the use of imagination and crea¬
tivity. Measurement of divergent thinking may be in terms of
absolute numbers of sensible answers, but other ways of assess¬
ing can include scores for fluency, originality, flexibility (i.e. a
number of different categories into which replies could be fitted)
and elaboration of responses.
2. In a number of studies the expression ‘foreign children’ was

used, which perhaps reflects the attitudes felt towards the child¬
ren of immigrants; the term was dropped later, presumably as a
result of changing attitudes.
3. Or, rather, ‘double semilingualism’, because the notion refers to

the bilingual’s two languages, both of them known only partially.
Chapter 7
Sociocultural aspects of bilingualism
7.1 Issues raised in the discussion of bilingualism
The controversy between proponents and opponents of bilingualism is

an old one. We can find statements on the effects of bilingualism made
as long ago as the beginning of the nineteenth century, and even earlier
(the distinction between bilingual education and bilingual upbringing
was not made at the time). The debate was often fuelled by people
who seem to us today to have been unqualified to pass judgement be¬
cause of their apparent lack of first-hand experience of bilingualism
and bilinguals. Yet the views of philosophers such as Fichte (1762-
1814) and Herder (1744-1803), educationalists like Jahn (1778-1852)
and Laurie (who published a series of famous lectures in 1890), philol¬
ogists such as Wilhelm von Humboldt (1767—1835) and, earlier this
century, Schmidt-Rohr (1932; 1936) and Weisgerber (1933), and per¬
haps even some psychologists and paediatricians (e.g. Sander 1930;
1934; Pichon 1936), have done much to discredit bilingualism.* The
present century has seen a good deal of polemical writing on the issue,
especially in the wake of extreme nationalism and fascism in various
parts of Europe and beyond. During the last thirty years, however, an
increasing amount of detailed investigation has led to a re-assessment
of some of the issues raised previously; with the application of more
modem methodologies based on empirical research, new insights have
been gained and a more balanced picture has emerged.
If we consider some of the views on bilingualism advanced over the
last 150 years or so it is possible to discern the main issues that have
attracted the attention of scholars. It will also become apparent that
opinion has swung full circle, at least among those who may be re-
VIEWS OF BILINGUALISM 137
garded as experts in the field. Nowadays bilingualism is no longer con¬

sidered to impair the intelligence of the individual (or the whole ethnic
group) or to be the cause of emotional problems. Some even say that,
on the contrary, it can contribute to enhanced intellectual abilities (see
Chapter 6). Bilingual education, for so long scorned and neglected, is
now seen by many as an attractive proposition. What has caused this
dramatic change?
The main point of departure in the debate on bilingualism has
usually been the assumption that it has some kind of influence on the
bilingual individual. Numerous studies have defended the theories of
either positive or negative effects of bilingualism on, for instance, cog¬
nitive development or character formation. In fact, it is not easy to find
anyone today, whether specialist, teacher or lay person, who does not
hold an opinion, often a strongly felt one, on the subject. But why
should bilingualism have a particular effect on the individual? No one
ever asks whether mo«olingualism exerts a specific influence on the
speaker - and in view of the fact that well over half the world’s popu¬
lation is bilingual this does not seem to be an unreasonable question to
ask. Why should bilingualism, any more than any other concrete skill
or ability (e.g. musical), carry any consequence for a person’s develop¬
ment and character? It is rarely bilingualism itself (on its own) that has
been proved to have had any measurable impact. More often than not,
it is the socio-psychological context of the bilingual individual that can
be shown to have a bearing on his or her evolution as an individual
and a member of society.
Bilinguals differ from monolinguals only in that they use two differ¬
ent linguistic codes (which in the case of child bilinguals have been
acquired largely in the same way as monolinguals acquired theirs), and
in that they may have access to two different cultures. Just like mono-
glots, they use language in order to communicate and socialize, i.e. in
order to function as members of a social group. Whether their tool is
one language or two, or a sign language, does not hy itself make any
difference.
7.2 Views of bilingualism
The idea of a close link between language, nationality and nation

which arose out of the Romantic movement, in the late eighteenth and
early nineteenth centuries, was to have far-reaching consequences for
138 SOCIOCULTURAL ASPECTS OF BILINGUALISM
bilingualism, as well as in other respects. It was seized upon by some

writers and politicians during the next few decades, who used it to
support their arguments in favour of the supremacy of the nation state.
A national language was seen as a sign of the nation’s unity and a
symbol of its identity. This kind of ideology paid scant attention to the
needs of linguistic minorities. The influential German educationalist
Friedrich Ludwig Jahn, best known today for his efforts to introduce
physical training as part of the curriculum, spoke out strongly against
bilingual education, claiming that it would retard the linguistic and
cognitive development of children. He, and others after him, argued
that a child can cope with only one mother tongue. The first language
was also said to impart a particular view of the world, so if the young
learner was confronted with two different visions or perspectives he
was likely to be tom between divergent Weltanschauungen and loyal¬
ties (Jahn 1808). At the end of the nineteenth century this view was
still widespread; for instance, the English educationalist Laurie said:
If it were possible for a child to live in two languages at once equally well,
so much the worse. His intellectual and spiritual growth would not thereby
be doubled, but halved. Unity of mind and character would have great diffi¬
culty in asserting itself in such circumstances.
(Laurie 1890: 15; quoted in Baker 1988; 10)
The claim that bilingualism causes intellectual retardation, linguistic

chaos and conflicting or split identity was still being made by various
writers in the first third of the twentieth century. Some, such as Weis-
gerber and Schmidt-Rohr for example, were directly influenced in their
thinking by the prevailing German nationalist ideology. But there were
others - linguists, doctors and sociologists from many different back¬
grounds — who, on the basis of bona fide research and without any
apparent ideological bias, wrote that bilinguahsm had a negative effect
on the development of the individual: left-handedness and stuttering, or
even, in certain cases, intellectual or moral inferiority and social margi-
nality were, in their view, possible consequences.
In his book Languages in Contact (1953) Weinreich included an
appendix entitled Effects of bilingualism on the individual’, in which
he surveyed the writings of some fifty authors on the subject. In the
majority of cases negative views had been expressed. This is not sur¬
prising, because it was clear (from Weinreich’s survey and from later
work by other linguists) that people’s feelings about bilingualism were
shaped by their opinions at the time on matters as diverse as intel-
VIEWS ON BILINGUALISM 139
ligence, language learning, education, the position of minorities, and

the status of one’s mother tongue as a world language.
For a long time many European states pursued centralist policies
which allowed for little expression of cultural or linguistic diversity.
This meant that members of linguistic minorities were obliged to learn
the language of the dominant majority in order to survive and to par¬
take in public life. The majority’s views on bilingualism were (and still
often are) influenced by their associations with, and prejudices about,
the minority or minorities existing in their midst. If the minority group
is suppressed, placed in a position of social and economic disadvantage
and made to feel culturally insecure, and if their language receives no
support, that society’s negative attitudes towards the members of the
minority group(s) extend to their behaviour, including their linguistic
behaviour - and from that point on to bilingualism itself. If this is the
case, it is easier to blame bilingualism for unsatisfactory school
achievement or low social status than to attribute shortcomings of this
kind to inequality of treatment for which the nation as a whole could
be held responsible. Eventually, some minority members end up having
a low opinion of themselves and, as a consequence, their bilingualism
may help to reinforce the idea that using two languages creates prob¬
lems for the individual. This is why bilingualism has been considered
by many immigrants as a sort of halfway house between monolingual-
ism in the ‘home’ language and monolingualism in the language of the
new country, and as a sign that total integration and acculturation have
not yet been achieved.
Bilingualism is sometimes regarded by minorities as a necessary
evil, representing a compromise between the need, on the one hand, to
hang on to their own language and identity and, on the other, to main¬
tain political and economic links with the majority. This may well
involve a conflict of interests - the desire for independence clashing
with economic and political improbabilities - and bilingualism might
then be seen as a symbol of such conflict. The fact is that, in some
parts of the world bilingualism has been (or is still) viewed with suspi¬
cion by members of both majorities and minorities. Eor the former it
may mean the threat of possible conspiracy, a code in which the mi¬
nority communicate in ways not understood by the majority. Eor the
latter, the minority aware of the unstable nature of their bilingualism, it
may represent the last stage before linguistic capitulation, i.e. shifting
towards monolingualism brought about by the decline of the minority
language and all it stands for.^
Attitudes towards minorities are slowly changing. The people of

many states are finding it increasingly easier to take a certain pride and
interest in their indigenous minorities. On the other hand, their atti¬
tudes towards some of their new minorities tend to be guarded, even
hostile at times. It is true that most European minorities today enjoy
more freedom and support for their language than fifty years ago, and
many bilingual services are available where previously there was none.
The reason for this is that no European country has failed to feel the
impact of the wave of post-war migration, greater social mobility, and
phenomena such as temporary work abroad and intermarriage, as well
as new initiatives in political, military and economic cooperation. All
of these have brought together, to a greater extent than in the past,
many different individuals, including members of minorities. As a res¬
ponse to social reality, views on bilingualism were bound to change.
The large number of publications, often using more reliable methods of
investigation than the old ones, that have challenged the negative senti¬
ments expressed by earlier writers have also done their share in
bringing about a better climate of opinion.
To a large extent, therefore, the discussion of bilingualism has cen¬
tred around social issues raised by the coexistence of majority and
minority groups. In this section, we shall look at three areas of re¬
search which have received special attention, all three connected with
the debate about the possible effects of bilingualism (as Grosjean
1985a: 469 put its, ‘the “effects” literature’): the psychological, the
linguistic and the cognitive consequences of being bilingual. They bear
strongly on questions of pedagogy and school curriculum; this lit¬
erature deals, then, with matters relating to the education of bilinguals
as well.
Much has been written about bilingual education; so much, indeed,
that this area of study has developed into a subject in its own right. In
1953 UNESCO published an historic document in which it declared
that every child had the right to receive primary education through the
medium of the vernacular. Discussion of the education of the children
of minorities, both indigenous and new ones, has been continuous ever
since. Unfortunately, the amount of knowledge gained has not (in Eu¬
rope especially) been matched by state provision in this field. The
subject of bilingual education is covered in detail, for instance, in
Swain and Lapkin (1982) and Genesee (1987) which deal with immer¬
sion programmes in Canada; Spolsky (1986), on multicultural and
multilingual educational issues in various parts of the world; and
PSYCHOLOGICAL ASPECTS 141
Husen and Opper (1983), Shapston and D’Oyley (1984), Cummins

(1984a), Churchill (1986), Byram (1986b), Boos-Nunning et al.
(1986), the OECD CERI Report (1987), Baker (1988), Skutnabb-Kan-
gas and Cummins (1988) and Byram and Leman (1989), which look
into a variety of aspects of the education of minorities, mainly within
the European context.
7.3 Psychological aspects
In this section we consider two areas. The first is of a psychosomatic

type, and it will be only touched upon so as to concentrate more on the
question of whether bilingualism can be said to have any detrimental
effect on the bilingual’s personality development.
7.3.1 Stuttering
It has been claimed that bilingual children are more likely to stutter
than monolingual ones. This claim was based on reports on the obser¬
vation and treatment of several instances of language disorder in which
bilingualism was deemed to have been a contributing cause. For in¬
stance, in 1937 the French doctors Pichon and Borel-Maisonny
published a book on stuttering (it was reprinted in 1964); 14 per cent
of the children in their study who stuttered used more than one lan¬
guage. Pichon (1936) had previously argued that bilingualism had
detrimental effects on children’s cognitive development, which perhaps
helps to explain the background to the later work. Investigations car¬
ried out later by Lebrun and Paradis in east Chicago indicate that the
symptoms described in the Pichon and Borel-Maisonny case histories
are also found in monolingual stutterers (Paradis and Lebrun 1984: 12
ff.). Another study undertaken in the 1930s (Travis et al. 1937) also
came to the conclusion that the incidence of stuttering was higher
among bilinguals (2.8 per cent) than for monolinguals (1.8 per cent).
Paradis and Lebrun deal with this claim, too: in their reassessment the
subjects are listed according to race and linguistic origin, and they find
that, if this is done, black English-speaking monolinguals and children
who could not speak English well are much more likely to have higher
scores for stuttering than the rest. They show, therefore, that any corre¬
lation between bilingualism and stuttering must be unreliable. They
suggest that stuttering is a neurotic symptom for which there may be
any number of psychosomatic or socially induced reasons; bilingual-
ism, however, has not been shown to be one of them, although stam¬
mering may of course be brought about in a bilingual child whose
social and educational experience is so devastating as to disturb his/her
psychological well-being.
7.3.2 Personality development

Much of nineteenth- and early twentieth-century thinking on education
was influenced by the conviction that the development of the person¬
ality of the child is shaped by the language and culture in which the
individual grows up. For some authors, e.g. Weisgerber (1929) and
Schmidt-Rohr (1932), this meant positing a close connection between a
particular language (say, German or English) and the development of a
German or English child. Naturally, such theoretical standpoints came
up against the problem of how to accommodate two languages and
cultures, in the case of the bilingual child, and the view that bilingual¬
ism would bring about ‘marginal men’ suffering from split identities
was voiced fairly frequently (see Goldberg 1941). But long before a
series of German authors had condemned bilingualism in the 1920s
and 1930s, writers in other parts of the world had been warning about
emotional problems caused by bilingualism (see Weinreich 1968: 117
ff. for a survey and some striking quotations).
The basis for many of these claims was anecdotal evidence in the
fomi of informal observations and comments by individuals; later more
systematic work was undertaken into phenomena of social alienation
and the possible detrimental effects of bilingualism on personality de¬
velopment among members of minorities and immigrant communities.
In 1943 Child carried out a study of second-generation Italian Ameri¬
cans, looking into their attitudes towards their dual heritage. He
concluded that his subjects had reached varying degrees of accultura¬
tion. He identified three different groups:
(1) those who identified strongly with American social and cultural
values;
(2) those who orientated themselves more strongly towards their Ita¬
lian background, rejecting everything American;
(3) those who refused to think in terms of these parameters.
As part of a wider study of Italian immigrants in Bedford, Great

Britain, Arturo Tosi (1983) also describes three distinct sets of attitudes
he observed among second-generation Italian Britons, classifying them

according to their reactions to social forces and to the pressures exerted
by their rather conservative Italian home background,
(1) People who displayed ‘apathetic’ reactions were in the majority.

They were those who could not cope with the pressures resulting from
the two cultural contexts in which they found themselves and therefore
yielded to the greater influence - that of the wider social environment
- trying to conform to British values while trying to suppress their
Italian background (but of course not being able to do so totally):
‘Here the passive attitude that has made him [the individual second-
generation immigrant] unable to sort out his own personal conflict also
prevents him from understanding the functions and roles of the two
opposing sets of values and behaviour’ (Tosi 1983: 116).
(2) The ‘in-group reaction’ is the name given by Tosi to the attitudes
displayed by those who identify strongly with the values of the Italian
community. He argues that the very limited facilities available to the
immigrant group he studied, and the lack of opportunity for more var¬
ied social interaction, were largely to blame for the youngsters’
apathetic and in-group attitudes towards the ‘host’ society.
(3) Only a small group of young people showed what he calls ‘rebel
reaction’, i.e. rebellion against the previous two extremes. In a con¬
scious effort to become bilingual and bicultural, they have refused to
retreat into the old culture (Italian) or to ‘sell out’ to the new one
(British). Again, Tosi emphasizes here the importance of wider social
opportunities, educational help and communication outside the host
community (i.e., contacts with Italy and Italian culture).
A series of studies on young people’s attitudes towards their mixed

parental background was carried out by Lambert and his associates in
the 1960s and 1970s. Their subjects were American and Canadian Eng¬
lish-French bilinguals, both children and adolescents. None of these
investigations, which were carefully matched with monolingual control
groups and controlled for factors such as socio-economic status and
linguistic standards, found any correlation between bilingualism and
adversely affected personality. Different patterns of attitudes could be
discerned, in line with the ones described by Child (1943) and later by
Tosi (1983): some people oriented themselves towards the English or
towards the French backgrounds, others did neither, and a third group
was successful in adapting to a mixed French and English sociocultural

environment (Aellen and Lambert 1969; Gardner and Lambert 1972).
The largest groups in Child (1943) and Tosi (1983) were undergoing
a shift from bilingualism towards monolingualism. But the assumption
that shedding a language speeds up acculturation or helps avoid culture
conflict is probably an erroneous one. For the individual who finds
himself/herself cut off from his/her first language it may be necessary,
or practical, to concentrate on his/her new language and culture. But
for a minority group a shift away from bilingualism may not solve the
social or psychological problems its members experience - on the con¬
trary, they may become aggravated. A Turkish community in, say, the
Federal Republic of Germany is likely to be considered foreign
whether they have lost their mother tongue or not. Therefore, to aban¬
don the home language in the hope of ridding oneself of the stigma of
being a member of an ‘immigrant group’ may be too high a price to
pay - and pointless, anyway.
The incidence of psychosomatic illness among migrant workers and
refugees in western Europe seems to be disproportionately high (Appel
and Muysken 1987). Their children are reported as showing a com¬
paratively frequent occurrence of behavioural disturbances, and they
are said to underachieve consistently at school. But there is no indica¬
tion that these problems are caused by these people’s bihngualism.
There are, on the other hand, plenty of reasons to think that they may
be the result of the diverse pressures they are exposed to, as a conse¬
quence of being the most stigmatized and underprivileged members of
society: the subjects are more likely to suffer from poor housing, un¬
employment and social marginalization than any other section of their
communities. It has been shown that many problems can be eased
when subtractive bilingualism is turned into additive bilingualism. This
is the case when the majority begins to show some appreciation of the
minority group’s culture and language, for instance by allocating radio
and television time to it and including its study in the school curricu¬
lum. Appel and Muysken (1987) report on research carried out in the
Netherlands on the social and emotional development of Turkish and
Moroccan children; those who took part in a transitional bilingual pro¬
gramme showed fewer problems than a comparable group of children
who had attended monolingual Dutch schools.
The two studies by Lambert and others mentioned earlier (Aellen
and Lambert 1969; Gardner and Lambert 1972), as well as the one
undertaken in cooperation with Tucker (Lambert and Tucker 1972), in-
volved settings of additive bilingualism. Subjects were all reported to

be happily settled in their English-French backgrounds in Canada and
the USA; many of them came from middle-class families which pro¬
vided strong support for bilingualism. Lambert was also closely
involved in a Montreal pilot project (known as the ‘St Lambert Experi¬
ment’), which aimed at making monolingual majority (English-
speaking) children bilingual in French and English. For this purpose, a
group of children who were monolingual in English were given their
initial schooling entirely in French. Their progress was carefully moni¬
tored and compared with that of a monolingual control group taught in
English (reported in Lambert and Tucker 1972). Although the medium
of instruction was French and the teachers tended to use only French
when speaking to the pupils, the latter could use their common LI in
the classroom and in the playground. This was in significant contrast
with the situation of the minority child who enters mainstream educa¬
tion in the L2. The children’s acquisition of French was very
successful, and the positive results led to the launch of national immer¬
sion programmes all over Canada. The main benefits (for the state) of
these programmes can be seen in the part they play in lessening lin¬
guistic prejudice towards the French-Canadian minority. In their 1972
publication Lambert and Tucker reported that a change in attitudes
could be observed in children who, after grades five and six, had
become functionally bilingual: these children were beginning to ident¬
ify with Canadian and European French culture, as well as with
English Canadian; they felt at ease in French social settings and ex¬
pressed positive feelings towards French-speaking people. The authors’
researchers therefore felt confident enough to voice their conviction
that biculturalism had been achieved with no loss of English identity.
Up to now these immersion programmes have involved only French
for English-speaking majority children. One must hope that eventually
the several other linguistic minorities of Canada (for example the spea¬
kers of Inuktitut, an Eskimo language spoken in northern Canada by
the Inuit people) will also be able to benefit from the additive biling¬
ualism afforded to its linguistic majority by means of bilingual
education programmes.
The vast majority of bilinguals in the world never have a choice
between staying monolingual or learning two languages. With the ex¬
ception of elite bilinguals and children from majorities undergoing
immersion programmes, bilingualism is usually a social or economic
necessity. It is with respect to these people that the phenomenon of
anomie has been discussed. The term was originally introduced by the
French sociologist Diirkheim (1858-1917), to describe the feelings of
rootlessness, social isolation and personal disorientation experienced
by those who were in the process of moving from one social class to
another. Applied to the bilingual (and also to people who, through in¬
tense motivation, become proficient in a foreign language), anomie
may denote ‘feelings of chagrin or regret as he loses ties in one group,
mixed with fearful anticipation of entering a relatively new group’
(Lambert, Just and Segalowitz 1970: 274).
Anomie may also result from pressures brought about by conflicting
cultural norms and loyalties felt by second-generation immigrants.
Both Child (1943) and Tosi (1983) mention the inability of many
young immigrants to make positive choices and to resolve the discor¬
dant demands made on them by the home and the wider communities —
what usually happens in practice is that they are not at all free to
choose. Baetens Beardsmore (1982) sees anomie as one of the more
important problems implicit in becoming bilingual. The person ‘who
tries to reconcile two widely divergent linguistic and cultural patterns
may find the inaccessible goals [of achieving balanced ambilingualism]
presented to him by his two environments leading to feelings of frus¬
tration’ (Baetens Beardsmore 1982: 127).
He argues both that anomie is more likely to result when the cultu¬
ral norms of the two communities are highly differentiated, that it
affects adults and adolescents more deeply than children, and that it is
more often found among individuals of low socio-economic status.
It appears that anomie is primarily a psychological phenomenon
brought about by certain sociocultural and socio-economic constella¬
tions. Whilst there is no necessary correlation between features of
bilingualism and factors of personal development (i.e. a bilingual’s per¬
sonality cannot be said to be pre-determined to evolve in any particular
direction), it does seem that at certain times particular bilingual groups
or individuals may be especially vulnerable to some kind of psycho¬
logical instability. As a result of a feeling of being pulled in different
directions and being unable to resolve the ensuing problems by turning
to either one language or culture, anomie may, in extreme cases, occur.
The question of Who am I?’ is of course asked by monolinguals and
bilinguals alike, but in an attempt to find an answer the latter has the
additional dimension of his/her biculturalism to contend with. There is
now a growing number of immigrant writers who have expressed their
search for identity in literary form. Some are intrigued by the pluralist
LINGUISTIC ASPECTS 147
nature of migration, a consequence of the fact that this century has

experienced migration - and loss of self - on an unprecedented scale.
An atmosphere of linguistic and cultural equilibrium, brought about by
parental and social support of both languages and cultures, can serve as
counterbalance, as it will facilitate a sound development of the psycho¬
logical make-up of the bilingual child. The most effective way in
which a state can contribute to the bicultural person’s well-being is by
providing a variety of types of educational and cultural facilities, so as
to suit the different needs of bilingual citizens.
In connection with the issue of the development of personal identity,
bilinguals themselves sometimes report having experienced problems
when asked by monolinguals about their national allegiance. Children,
in particular, often find such questions embarrassing. They do not nor¬
mally identify with large groups, and the concept of nationality is not
really understood by them, apart from some perception of a link with a
particular language or person. They may say that they are a little of
this and a little of the other, referring to both the parents’ ethnic
group(s) and the country where they live, or they may insist that they
are just the same as other children in their environment. What they
seem to resent most is being singled out on the basis of the languages
they speak. In one of the chapters (entitled ‘So I think I’ll stay half¬
way’) of her book Many Voices, Jane Miller (1983) transcribes her
interviews with some young bilingual girls in London who talked
about the frustrating feeling of not being quite accepted as natives of
either of the two cultures they had access to; for example, Maria, a girl
of Spanish parents, then aged seventeen, said: ‘When I go back [to
Spain] they call me, “Oh look, the English girl”. Oh, I get annoyed.
And when I come here, “Oh look, Spanish girl”. So I think I’ll stay
halfway. . . but you have to sort of sacrifice yourself’ (Miller 1983:
24).
Similarly, one of my children was quite indignant (at the age of 9:5)
when he found that, while staying with friends in Germany, some
children at the school he was attending noticed at once from his
spoken German that he was not from the region. They asked him
where he came from, and from that moment they called him ‘der Eng¬
lander’. He seemed frustrated that, in spite of his trilingual upbringing
(which led him to believe that he was German, Spanish and English),
he was nevertheless considered a foreigner. Such incidents bring home
the message that raising children bilingually requires also that parents
point out to them the various ways in which they are perceived to be
different from monolinguals, so as to help them define and sort out

their identities.
Bilingualism does not predispose the subject towards psychological
problems, but for the individual bilingual it may require a more con¬
scious effort to establish his/her identity. The socialization process of
children entails modelling their identity on that of the community. In
the case of the bilingual child two models present themselves, thus
making socialization more complex.
7.4 Linguistic aspects
Before the spate of research carried out in the 1960s into various as¬
pects of bilingualism, minorities and education, one view that was
commonly held among lay people was that bilingualism had a det¬
rimental effect on the child’s linguistic skills. The idea that a second
language is acquired at the cost of the first, and that therefore neither
can ever be mastered fully, was widespread. The observed markers of
bilingual speech were regarded negatively, and they offended monolin¬
gual speakers’ ideals pf purity af language. Little thought, if indeed
any at all, was given to the notion of bilingual language competence as
being different from the monolingual’s,; nor was it ever suggested, at
that time, that mixed language could be anything but a sign of imper-
fecTlinguistic knowledge. ..
7.4.1 Assessing bilinguals

The early literature on bilinguals (reports on both individuals and
groups) varies greatly in methodology and therefore in reliability. It
was often the case that the bilinguals who were the object of attention
were compared with monolinguals without any attempt being made at
controlling the variables that today are known to play an important
part, such as the children’s socio-economic and educational back¬
grounds, and their levels of intelligence and motivation. Also, the
research tools available for measuring and comparing linguistic ability
and performance were sometimes crude, consisting for instance of
simple language tests administered to both bilingual and monolingual
groups in the common tongue, without paying any heed to such issues
as which was the weaker/stronger language in the case of the bilingual
speakers. But considerable progress has been made in the last few
years in the areas of assessing bilinguals and measuring their language
LINGUISTIC ASPECTS 149
skills by making reference to (1) those features which are evident in

monolinguals and (2) those which are characteristic of bilinguals only.
7.4.2 Linguistic competence as the object of assessment

Bilingualism is an unstable phenomenon and, as we know, bilinguals
differ a great deal from one another in the degree of proficiency they
achieve and the extent to which they maintain their languages. Bilin¬
gual children usually gain full communicative competence in at least
one of their languages; this competence includes both the internalized
knowledge of a particular linguistic code and the ability to use it (more
or less) effectively in acts of communication. Carolyn Kessler (1984)
establishes a theoretical framework for communicative competence in
which she identifies four areas of knowledge and skill. Taken together,
they provide a breakdown of what is involved in the process of lan¬
guage acquisition.
(1) Grammatical (or linguistic) competence refers to the mastery of

the linguistic code.
(2) Sociolinguistic competence applies to the acquisition of the
sociocultural rules that define the appropriate use of language in
different social contexts, e.g. children have to learn where, when
and to whom they can(not) use informal speech/styles.
(3) Discourse competence is concerned with learning to combine ut¬
terances into meaningful entities; it entails, for instance, the
knowledge of how to introduce or close a topic, and turn-taking
conventions.
(4) Strategic competence refers to the pragmatic strategies a native
speaker resorts to when communication breaks down and (s)he
needs to compensate for a failure in language performance: para¬
phrase, circumlocution, repetition and avoidance are examples of
such strategies, which may be used for ‘survival’ or for rhetorical
purposes.
Developing certain aspects of communicative competence can be a

long process for monolingual children. There are bound to be at least
some similarities between the acquisition of competence in Language
A and Language B, in the case of those acquiring skills in two or more
languages, particularly if the cultures involved are not too diverse. But
there may be wide variation in the extent to which these different as¬
pects of competence are developed. Also, as Kessler (1984: 30) points
out, there is a linguistic dimension to consider: ‘the acquisition of com¬

municative competence in two languages must further take into
account the interaction between two language systems’.
A number of interrelated factors will impinge upon the simultaneous
or sequential acquisition of communicative competence in two lan¬
guages.
(1) The linguistic environment will determine the nature and degree of
language input. For instance, language used in the home and among
friends tends to be highly context-bound, and speakers can depend on
paralinguistic features such as mimicry, gestures and non-linguistic
sounds. Language at school, on the other hand, differs increasingly
from the ordinary use outside as the child progresses through the edu¬
cational system: the content is constantly being renewed, and the
language is often (whenever issues unrelated to the immediate class¬
room environment are discussed) highly decontextualized. In addition,
language is employed frequently for developing new skills, such as
reading and writing, and cognitive processes - for example, calculation
or multiplication.
(2) The relationships between the child and, on one hand, the home
and, on the other, the teachers, school friends and members of the peer
group, will influence the kind of social interaction and communication
situations experienced by the child.
(3) 'Affective variables, such as the child’s emotional bonds with
various speakers of her two languages, or the perception of prevailing
attitudes towards both her languages and cultures, may also be of cru-
eial importance, as will a large number of personal characteristics.
(4) The need to use the LI and the L2 will result from a combination
of factors (1), (2) and (3), to name just some.
Any change in the environment can affect a bilingual person’s com¬

municative needs. Contact with one of the languages may be
interrupted, or increased, and as a result the pattern of her bilingualism
will be altered: passive may turn into active bilingualism, or vice versa.
There are only a few studies that deal with recessive bilingualism, and it
is not yet clear to what extent a language can be fully forgotten, or
retrieved, once left behind. Burling (1959) and Murrell (1966) report
on the rapid loss of a second language in early childhood, once all
association with it had ceased. In the case described by Spndergaard
(1981), language loss or, as he puts it, the ‘decline of bilingualism’ was
probably due to a combination of lack of continuity of contact and
LINGUISTIC ASPKCTS 151
emotional factors. Berman (1979) writes on her young daughter’s tem¬

porary loss and subsequent recovery of Hebrew, while maintaining
English, the language of the parents. Older bilinguals are less likely to
descend into total unavailability of one of their languages, as in most
ca.ses they will retain at least some rudimentary form of passive bilin¬
gualism on which they can rebuild or ‘re.stmcture’ (Grosjean 1985a)
their language competence. So, the detailed configuration of a bilin¬
gual’s language is quite likely to vary over the period of his/her life.
Competence in either language may change and shift in order to adapt
to new .situations, new environments, new communication needs; but,
of course, one’s general communicative competence is not radically
altered by this.
As far as communicative competence and language learning and
forgetting are concerned, Gro.sjean challenges bilingual research to ex¬
plore new avenues: rather than trying to measure levels of attained
grammatical competence in each language taken individually and out
of context, one should endeavour to find out, he believes, ‘how the
human communicator adjusts and uses one, two or more languages -
separately or together - to maintain a necessary level of communica¬
tive competence’ (Grosjean 1985a: 473).
When investigating bilingualism, then, care should be taken to
choose sample groups who are matched according to the kind of bi¬
lingualism they pos.sess (whether active or passive, for instance), in
relation to the learning situation under which they became bilingual,
e.g. whether theirs was subtractive or additive bilinguali.sm, and also
with respect to their bilingual competence.
7.4.3 Measuring bilingual language skills

The traditional method of gauging bilinguali.sm was to assess individ¬
ual language skills and proficiency in each language by, for instance,
testing bilinguals’ understanding and production of certain lexical or
grammatical items, in either spoken or written form, or by measuring
their reaction to various linguistic clues (in terms of speed, or correct¬
ness), or by asking them to carry out linguistic operations (such as
filling in blanks, retelling a story, completing sentences etc.). A num¬
ber of different techniques have been devised for appraising bilinguals,
with a view to arriving at valid ways of comparing them with monolin-
guals. For example, records of spontaneous speech can be analysed for
a number of variables, such as sentence length and complexity or rich-
ness of vocabulary; the recordings are often played back to monoglots,

who pass judgement on the language (without necessarily knowing that
the speakers are bilingual); a wide range of language tests can be used
to assess mastery of syntax and lexis. There are also procedures for
finding out which of the bilingual person’s languages is the dominant
one. They include word-association tests, true/false statements and
presenting lists with cognate pairs (e.g. French ‘plume’, Spanish
pluma’) or words with identical form and meaning in the two lan¬
guages (e.g. German ‘real’, English ‘real’) but of course different
pronunciation, to be read out (the language to which the pronunciation
chosen most often belongs is considered to be the stronger one). Other
ways of assessing relative proficiency involve various translation exer¬
cises, responses to spoken/written stimuli, or verbal/pictorial
instructions. Many of these tests are carried out under laboratory con¬
ditions, and the response time is recorded; the higher speed of reaction
is then attributed to the dominant language.
7.4.4 The problems of assessing bilingual proficiency

Measuring bilingualism remains notoriously difficult, and so far none
of the methods and techniques used has been found to be generally
satisfactory. It is not easy to construct tests that are entirely valid if the
appropriateness of the setting within which they are administered is
taken into account. Any sampling and matching procedures with con¬
trol groups raise the question of comparability. Another problem area
concerns the nature of the linguistic means employed. For instance, the
language used for a given test may be related to a topic, or couched in
a style, unknown to the bilingual person being assessed; it may not
accurately reflect his/her social or cultural experience; or it may re¬
quire the use of skills (e.g. reading or writing) not normally used by
the subject in the language being evaluated. Another danger is the
failure to take account of the fact that a bilingual’s language com¬
petence (which draws on the knowledge of two languages) is different
from that of a monolingual. If the test battery is designed in such a
way that it is to be used with monolinguals as well as bilinguals, then
it is unlikely to contain the kind of language many bilinguals are most
familiar with, namely language incorporating speech markers (such as
mixes and switches) that are frequently present in bilingual talk. Also,
it may provide few opportunities to use such bilingual language as the
subject would normally employ.
NOTES 153
More generally, with reference to the interpretation of bilingual data

it is imperative to approach the task from different angles (see
Hoffmann forthcoming for examples). Grosjean (1985a) highlights the
importance of recording also the bilingual speech modes. These vary
according to the situation in which the speaker/hearer finds himself,
ranging from interaction with monolinguals in either of the two lan¬
guages to talking with bilinguals who share his/her two codes. What
might look like interference from one language on the other may well
be intentional borrowing or code-switching in the bilingual mode. As
Grosjean laments: ‘Rare are the bilingual corpora that clearly indicate
the speech mode the bilinguals were in when their speech was re¬
corded; as a consequence, many unfounded claims are made about the
bilingual’s knowledge of his or her languages’ (Grosjean 1985a: 474).
Formal tests tend to stress linguistic form, as in order to answer any
given item it is necessary to produce or recognize specific linguistic
units. However, such responses may not reveal the bilingual’s ability to
communicate. Thus, in many cases, the results obtained can only allow
the formulation of tentative statements about the bilinguals’ partial
proficiency in each language, but not about his or her full bilingual
communicative competence.
Notes
1. The case of Humboldt is remarkable. At least two of his children
grew up bilingually while they were staying with their father in
Italy; in letters to his wife (see some extracts in Porsche 1983: 68
ff.) he comments on their language use, in which Italian was be¬
coming increasingly dominant, and it seems that he saw nothing
unusual in bilingual acquisition. It is curious that Humboldt’s
theories on the structure and meaning of language were later
used for the idealization of monolingualism.
2. Several instances from various chapters of European history

could be quoted. One clear example is provided in the 1847 ‘Re¬
port of the Royal Commission of Inquiry into the State of
Education in Wales’. The three Commissioners, all Englishmen,
found that schools in Wales were inadequate (which is not sur¬
prising, as many were so in most of Britain at the time).
However, they blamed the Welsh language for the failures of
education in the Principality, calling it ‘a manifold barrier to the
moral progress and commercial prosperity of the people’ (cited
in Baker 1985: 42). As far as the administration of justice was

concerned, their judgement was even harsher, as they contended
that the ‘evil’ of the Welsh language was obviously great, since
it distorted the truth, favoured feud and abetted perjury. In a dif¬
ferent century and another part of Europe, the careful avoidance
of any reference to bilingualism in the wording of Catalonia’s
many official pronouncements on language use in the region can
be taken as evidence of many Spaniards’ suspicion of the sub¬
ject. Castilian Spanish is the national language, and therefore
every citizen (according to the 1978 Spanish Constitution) has
the duty to learn it. Catalonia promotes Catalan as its own lan¬
guage - so bilingualism is the result. But no mention is ever
made of the word: Catalan, not bilingualism, is officially pro¬
moted (see section 13.2 for a fuller discussion).
PART II
SOCIOLINGUISTIC ASPECTS OF
BILINGUALISM
>»ocr«ftTV»tAi ga jwouAti5»lf
uiite6ft>eE4, ♦‘atJw* aji <twy o>ntei».! i

U' -- •-«; ‘ey\r ■ >r
a tlinitcmte. TO
I^Lît ftsiU-’y p'rj
\>f r k‘'BC jyi? m iti; <■ t;!* v
^nanv ^ttnai oa ukr jb 0k: region c«n
■fx* #' !?vi(li3Ka of oMiDy Sfvudt(n3)i' mpickvi of ibe <u^ ^
iject 'Cwabip^ ,%>afusii.ii iwtiomj ial\ûi»^ iucmI thax^^
zycty, ciib^ f*uwi«ia^ tr. jSvi l^JlS SpftOHb.CoiaiiUttiioft) hau |
«iniy ^ kM«rd H: Cmtha ^ )<'. awn t«ar
gunge - sa ite^ the r^Ji. fi«t m incniioo i> ncr
nt3d9 t.l the \W6rd- Caial.^. iioi uthm, t»
secdoR I>tfuller diûMaw '
■ ’"11
. 'li"' '■
I -.
Chapter 8
Societal multilingualism
8.1 Multilingualism in society
There is no reason to believe that monolingualism is the normal state

of affairs in human society. In fact, bilingualism is more widespread,
since more than half the world’s population can claim to be bilingual.
At the societal level multilingualism is quite common, particularly in
the continents of Africa and Asia. Bilingualism in the individual results
from a person’s contact with two (or more) languages. Multilingualism
comes about when speakers of different languages are brought together
within the same political entity. In the history of the world, the organ¬
ization of human society into states has rarely followed ethnic or
linguistic groupings.
Most countries display considerable linguistic diversity. Political
frontiers have proved to be much less stable than linguistic ones, and
each border change was likely to bring with it a change in linguistic
patterns. Many governments, particularly those of western Europe,
have chosen to ignore the language diversity within their frontiers, or
they have promoted the interests of the mother tongue of the dominant
elite. The philosophical underpinning for this attitude was the ideal of
‘one nation-one language’, which became fashionable in Europe well
over a hundred years ago. There is a history of suppression of minority
languages, for instance in Britain, France and Spain, which goes back
several centuries. However, neither indifference nor negative attitudes
towards non-dominant languages have resulted in general societal
monolingualism, although there is a threat to the survival of some
lesser-used languages. In what follows, the main focus will be on the
factors that have contributed to the emergence of multilingualism in
society.
158 SOCIETAL MULTILINGUALISM
8.2 Factors contributing towards societal multilingualism
The quest for political power and economic influence invariably in¬
duces language contact. By adopting first a diachronic and then a
synchronic stance, it is possible to see how a variety of factors all lead
to the same phenomenon, language contact, which often takes the form
of the spread of one tongue and its eventual dominance over another,
or several others.
8.2.1 Historical factors

(1) Military conquests, occupation, secession and annexation These are
some of the oldest ways of spreading language. Greek, Latin and
Arabic reached many parts of the world by these means in earlier
times, and in some territories they were the instrument of radical politi¬
cal and administrative changes; but once the political dominance
waned, the influence of the languages diminished as well, although in
varying degrees and at different speeds in the various parts of the
world where they left a mark. The Roman occupation left few direct
traces on the older languages of Britain (although Latin of course made
considerable impact on English later), but it contributed quite exten¬
sively to the lexicon of German, and it exercised a profound influence
on the development of French, Spanish, Portuguese, Catalan and
others.
(2) Political marriages and succession arrangements often brought

about changes in the linguistic make-up of particular areas. The mar¬
riage in 1469 of Isabel of Castile and Ferdinand of Aragon marked the
beginning of the downfall of Catalan and the dominance of Castilian
Spanish in Catalonia. Another marriage, that of Mary of Burgundy to
Maximilian of Austria in 1477, eventually brought the Low Countries
under Spanish rule (Charles V inherited the Burgundian territories
from his paternal grandmother, as well as the kingdom of Spain from
his mother, the daughter of Isabel and Ferdinand), as a result of which
Dutch and Flemish speakers found themselves governed, from 1519 to
1648, by a Spanish-speaking administration. And when the German¬
speaking House of Glucksburg acceded in 1733 to the Danish throne,
German became the language of the Danish nobility and higher eche¬
lons of society. However, neither Spanish in the Low Countries nor
German in Denmark became permanently established, and such bilin¬
gualism as had existed soon reverted to monolingualism.
FACTORS CONTRIBUTING TOWARDS SOCIETAL MULTILINGUALISM 159
Border changes often came about as a result of treaties following

large-scale armed conflict. Such arrangements sometimes resulted in
longer-lasting multilingualism. Examples are Italian-speaking South
Tyrol (the districts of Bolzano and Trento), on the border between Italy
and Austria, or the German-speaking parts of Belgium (Eupen and
Malmedy). Both of these present-day linguistic minorities were the
outcome of territorial changes stemming from the Treaty of Versailles
in 1919, after the Eirst World War.
(3) Colonization, mainly from the sixteenth century onwards, often on

a large scale, of vast areas in the Americas, Africa and Asia, was
fuelled by the political and economic ambitions of the colonial powers.
Because of its forceful nature, it can also (and perhaps more appropri¬
ately) be called imperialism, although, as Fasold (1984; 10) points out,
this is a ‘loaded word’. Colonization involved a relatively small num¬
ber of people taking control of territories inhabited by other peoples,
exerting political power and controlling economic development. The
colonial (or imperialist) groups, mainly the British, French, Spanish,
Portuguese and Dutch, adopted different policies with respect to the
administration of their colonies; but all of them introduced their own
languages in the areas under their influence. Some powers were more
concerned than others about spreading their languages. The British and
the Dutch took an essentially pragmatic position, one of their prime
objectives being the smooth running of the colonies; their languages
were used for purposes of administration and higher-level economic
transactions. English also became the language of education, usually at
the secondary level and beyond. For the Spanish and Portuguese, col¬
onization also meant the spread of the Christian faith, so territorial
conquests were followed (and in some cases preceded) by missionary
efforts - and it was part of the missionaries’ duties to teach the lan¬
guages alongside the Gospel. French colonialism, too, was
accompanied by a sense of mission, although one of a more cultural
than religious nature: it was part of the ‘mission civilisatrice’ to dis¬
seminate a knowledge of French and of the ideas of metropolitan
France, and therefore in the French colonies most aspects of public life
were organized according to French models. The French language,
consequently, was promoted in order to achieve ‘civilizing’ goals.
(4) Migration and immigration have taken place throughout the history
of mankind. Migrant groups have often become assimilated into main-
Stream society after some generations, as has usually been the case in
countries of immigration such as the USA, Canada and Australia (it is
not the case, however, that every single immigrant group or individual
has become assimilated in these countries). The type of large-scale mi¬
gration that assumes the form of colonization has often resulted in a
noticeable spread of the language of the settlers. For instance, the col¬
onization of eastern Europe by German speakers in the thirteenth and
fourteenth centuries contributed considerably towards the geographical
expansion of this language. But whereas at the beginning of the twen¬
tieth century German was still widespread in eastern Europe, changes
in the political fortunes of Germany since then have led to a much-re¬
duced presence of the language in that part of the world today. On the
other hand, migration by a sizeable group of Germans, united by relig¬
ious belief and the fear of persecution, to Pennsylvania brought a
particular variety of an eighteenth century south-west German dialect
to the United States, where it is still used by some of the descendants
of the original immigrants to this area. The language of these people,
the Amish, is sometimes referred to as ‘Pennsylvania Dutch’. It is
marked by heavy influence of English on all linguistic levels.
(5) Federation, whether of a voluntary or forced nature, has also con¬

tributed towards societal multilingualism. Switzerland and Belgium
have constituted themselves into federations of their own free will, and
in both countries the language groups that came together have
achieved official status for their languages, although it took a long
struggle for Flemish to achieve equal status with French in Belgium
and, in the case of Switzerland, the smallest community, the Romansch
speakers, were not successful until 1938. Forced federation often came
about when European colonial territories in Asia or Africa were
brought together under a single administration. After independence
some federations remained united, but in other cases some of the feder¬
ated groups concerned tried to secede from the larger body:
Bangladesh is a country that attained secession (from Pakistan), where¬
as Biafra (from Nigeria) represents an unsuccessful attempt.
8.2.2 Contemporary factors

(1) Neocolonialism Most former colonies are independent states today;
but imperialism has left many legacies in the linguistic field. Neocolo¬
nialism is characterized by economic (often coupled with political)
dependency and by the maintenance of the old colonial language. Lin-
FACTORS CONTRIBUTING TOWARDS SOCIETAL MULTILINGUALISM 161
guistic imperialism can therefore be seen as a consequence of econ¬

omic domination. For reasons of simple convenience, or on account of
financial stringencies, and in some cases because of the lack of a com¬
monly agreed alternative, many states have kept the old colonial
language as either the sole or joint official language. Many states no¬
wadays find themselves increasingly dependent, economically as well
as for purposes of obtaining the military equipment that they want,
upon larger countries - notably the USSR, the USA, Australia, China
and France - which also contributes to spreading the currency and use
of the respective languages. Tanzania was able to install Swahili as its
national language, but neighbouring Kenya was not - instead, it kept
English. India’s 1947 Constitution selected Hindi as the national lan¬
guage, and the intention was that it would gradually replace English;
however, the opposition of many Indians, particularly those in the
southern regions, forced an amendment to the Constitution in 1965
which allowed the continued use of English, alongside Hindi, for all
official purposes.
(2) Present-day immigration and migration of labour are continuing to

disseminate many languages across the world, thus contributing to cre¬
ative language contact, even in areas whose inhabitants have for a long
time considered themselves monolingual. The inhabitants of Central
and South America and the Caribbean have migrated to North America
in considerable numbers, especially in the latter half of the twentieth
century, and they have not been assimilated into the host society as
quickly as were earlier waves of immigrants. The new immigrants took
their own languages (mostly Spanish) with them; they have often
become bilingual, and they show no sign of giving up their mother
tongue. Britain, France, Belgium and the Netherlands have seen a
steady influx of people from their former colonies who, in many in¬
stances, have formed their own communities and have been successful
in maintaining their languages. The post-war economic boom in many
European countries attracted a large labour force from Mediterranean
countries. Some Scandinavian and certain EC countries have also ex¬
perienced the arrival of many political refugees from the Middle East
and Far East. It is too early to be able to say to what extent these
various groups of people will acquire the language of their new host
country and for how long bilingualism will remain a predominant fea¬
ture among them. At the present time, however, bilingualism is
certainly experienced as both an individual and a group phenomenon
among immigrant groups of various kinds, each language fulfilling

specific functions.
(3) Language promotion refers to government decisions reflected in

(more or less) well-defined language policies. This factor can contrib¬
ute to the spread of multilingualism quite considerably. The bluntest
kind of language policy is the imposition of one official language in a
particular area by suppressing the local one. This has occurred fre¬
quently in Europe, for example in the cases of the Basques, along with
the Gallegos and the Catalans in Spain, the Corsicans and the Bretons
in France, and the Gaelic and Welsh in the British Isles. Policies of
promoting Castilian Spanish, French and English, against the wishes of
local people in the respective areas, as the sole codes to be used in
administration and education have in the past proved very effective.
Language policies can also be of a more acceptable character, as the
promotion of modem Hebrew in Israel has shown. In some countries
where there is rich linguistic diversity and the continued use of the
colonial language is no longer desired, governments have actively en¬
couraged the development and use of certain pidgins, creoles and local
languages by declaring them official, or even national, languages. This
was the case with Bahasa Indonesia and Bahasa Malaysia in the
respective areas. Language promotion sometimes takes the form of lan¬
guage recovery or language replacement programmes: a language
whose use has diminished is promoted with the aim of halting its de¬
cline or enabling speakers to use it as an expression of their national
identity. The gradual reintroduction of Catalan as the official language
used in all walks of public life in Catalonia can serve as an example,
as can the efforts of the Greenlandic government, after independence
from Denmark in 1979, to establish their own autochthonous language
in administration and education. Whatever form language promotion
takes, it can be successful (that is, bring multilingualism in its wake)
only if it receives widespread popular support. It is heavily dependent
on positive attitudes, as otherwide it is unlikely to be allocated suffi¬
cient funding or receive public response.
(4) Internationalization The conditions of life in today’s world, includ¬

ing the mobility of labour across linguistic frontiers, require many of
us to deal with people from different language backgrounds. Interna¬
tional cooperation and communication needs have brought about the
advancement of certain languages, especially English. In many coun-
PATTERNS OF SOCIETAL MULTILINGUALISM 163
tries, foreign-language teaching in schools and in higher education has

contributed to producing many people with some kind of bilingual
skills. In some instances, whole communities and international organiz¬
ations (such as the agencies of the European Community) are actively
promoting language learning. Such endeavours tend to lead to individ¬
ual rather than societal bilingualism, but they can also give rise to a
sort of temporarily bilingual group, as happens, for example, when
scientists from different countries come together for an international
meeting and agree on the use of a common conference language.
Multilingualism, then, is the result of a variety of factors, only some
of which have been indicated above. Language contact can take differ¬
ent forms and follow diverse routes, leading to multilingualism of a
more transient or more permanent nature. The factors that have initially
caused language contact are not necessarily those which maintain
multilingualism.
8.3 Patterns of societal multilingualism
The pattern of language use that emerges in a particular multilingual

community depends, in part, on (1) the factors which contributed to¬
wards language contact in the first place; (2) the forces that determine
the extent to which the languages involved are now used; and (3) the
functions that each language has been allocated, whether by central
governments, local authorities or individuals.
8.3.1 Determinants of multilingual patterns

(I) The types of pressure under which language groups find them¬
selves are often reflected in the speech behaviour of their speakers. A
linguistic minority group can find itself politically suppressed by a
dominant elite group which may (but need not) be in a numerical ma¬
jority. In present-day Europe economic pressure is a more subtle form
(than political coercion) of ensuring one group’s dependence upon an¬
other. If a community that constitutes a linguistic minority wishes to
share in a country’s economic progress, its members will have to ad¬
just linguistically, i.e. adopt the language of the group that holds the
political and economic reins. The need which they perceive to learn the
language of the dominant group, both for reasons of educational and
social advancement and in order to share fully in the country’s public
life, can be considered a form of cultural pressure. Whatever kind of
pressure is involved, the group which is affected by it is more likely to

become bilingual than the one which exerts it.
(2) The distribution of speakers of different languages within the same

country produces varying patterns of multilingualism. In terms of geo¬
graphical distribution, there are few isolated communities that have
avoided the need to become bilingual. For a linguistic minority to re¬
main monolingual, the people concerned have to be cut off from the
mainstream communication network of the country in which they live.
Conversely, if speakers of minority languages are in close contact with
those of another language, as happens in the case of many migrants in
Europe’s larger towns and urban areas, the need to become bilingual
becomes paramount. But when the numbers of migrants are so large,
and their settlement so concentrated, that it amounts to ghettoization,
some of the minority members can remain monolingual and rely on
others of their group to act as intermediaries. In some minority areas,
such as the present-day Catalan-speaking one (which extends to both
sides of the eastern French-Spanish border), we find both a geographi¬
cal and a demographic concentration of local-language speakers in one
part (in this case Catalonia itself), whereas in another (on the French
side) speakers are more dispersed and societal multilingualism is less
widespread.
far as social distribution is concerned, there are bilingual groups
who belong to the social elite of a country and those who do not. ‘Elite
bilingualism’ was common, for instance, within the Russian and Ger¬
man aristocracies in days gone by, when Erench was used in their
social and political dealings. This kind of bilingualism was also wide¬
spread in colonial territories throughout the world at one time. ‘Popular
bilingualism , on the other hand, is associated with linguistic minorities
for whom acquiring the language of the dominant group has become a
question of survival. Most linguistic minorities in Europe today fall
into this category.
8.3.2 Basic types of societal multilingualism

No two countries display identical patterns of multilingualism. Lan¬
guage groups differ from one another, firstly, because different sets of
factors are brought to bear on each community and, secondly, because
the reaction of each individual member of society to those factors will
be unique. Yet it is possible to discern some basic types of multilin¬
gualism.
The term horizontal bilingualism (Pohl 1965) describes the situation

in countries where different languages enjoy equal status. It is not easy
to find examples that exactly fit this category, because the requirement
of ‘equal status’ is seldom met. Italian, Romansch, French and German
have equal official status in Switzerland, but Italian and Romansch are
less widely, and less frequently, used than the other two, and therefore
they can be said to have been attributed lower social prestige. Flemish
in Belgium, and French in Canada, are in a similar position. Unless,
therefore, ‘equal status’ is defined broadly as meaning ‘enjoying offi¬
cial or co-official status’, the concept is not useful.
It is probably better to distinguish between multilingual countries in
which the principle of territorial monolingualism applies and those
which have opted for territorial multilingualism. The former means
that, in each language area of a state, linguistic provision is offered in
only one language. Thus most inhabitants of French-, German- and
Italian-speaking Switzerland can, if they so wish, remain monolingual,
just as the Walloons in French-speaking Belgium often do and the
Flemish could (though many do not). In border areas it is more com¬
mon to find widespread bilingualism, i.e. territorial multilingualism, as
for instance in the Swiss towns of Biel/Bienne or Freiburg/Fribourg
and in many towns and villages near the frontier in French- and Ger¬
man-speaking Belgium. In countries with official territorial
multilingualism one finds that both languages are used side by side. In
western Europe this occurs only in minority areas where the minority
language is officially recognized, such as Finland and Catalonia, some
parts of Wales, and Friesland in the Netherlands; it is also the reality of
the officially Romansch-speaking Swiss canton of Graubunden, where
German and Italian are widely used. In other parts of the world this
kind of language organization is more common, e.g. in India, South
Africa, Singapore and Hong Kong.
The type of multilingualism most frequently found in western Eu¬
rope is characterized by the social inequality of the languages
concerned. The states of western Europe tend to contain large monolin¬
gual groups presiding over linguistic minorities lacking both official
recognition and language rights. Bilingualism is widespread among the
members of these minorities but rarely encountered among the ma¬
jority group, who are less likely to feel the need to learn the minority
language.
8.3.3 Diglossia
In all speech communities it is common to find different social and
regional dialects being used in ways that reflect a formal as well as a
functional separation of language varieties. In western Europe the
standard variety tends to be preferred for use in the media, on formal
public occasions and for wider communication beyond the region. It is
usually one of the objectives of the education system to impart famil¬
iarity with the standard form of the language to children who may or
may not have already had access to it in the home environment. A
person who habitually uses a speech variety which closely resembles
the standard form does not need to know any other dialect, but clearly
this does not hold for the dialect speaker.
There are, however, speech communities where all speakers need to
know at least two varieties because each language form is associated
with a specific set of social functions. This language situation has
become known as diglossia. Pohl (1965) calls it vertical bilingualism,
since the two varieties exist within the same speaker. Ferguson’s
(1959) article, entitled Diglossia’, has become a classic study of the
phenomenon. The definition that he gives is:
Diglossia is a relatively stable language situation in which in addition to

the primary dialect of the language, which may include a standard or re¬
gional standard, there is a very divergent, highly codified, often
grammatically more complex, super-posed variety, the vehicle of a large
and respected body of literature, heir of an earlier period or another speech
community, which is learned largely by formal education and is used for
most written purposes, but is not used in any sector of the community for
ordinary conversation.
(Ferguson 1959: 336).
The most important feature of diglossia is the functional specializa¬

tion of two varieties of the same language. The High variety (H),
Ferguson suggests, is typically used for sermons (in church or mos¬
que), formal speeches and public lectures, news broadcasts, in official
documents and written communication, most books and newspapers,
and in poetry. The Low variety (L), on the other hand, is the usual
medium for less formal situations, for purposes such as conversation
with family, friends and colleagues, instructions to waiters, servants
and workmen, informal radio and television programmes (e.g. ‘soaps’),
captions in political cartoons, and in personal letters and folk literature!
Both the H and the L varieties are used for oral and written purposes.
Ferguson discusses diglossia under nine rubrics, which include both

linguistic (e.g. grammar, lexicon, phonology) and sociolinguistic (e.g.
prestige, function, stability) entries, and he uses examples from four
speech communities and their respective languages: Arabic (H = classi¬
cal Arabic, L = local varieties of individual Arab countries). Modem
Greek (H = katharevusa or katharevousa, a kind of puristic variety con¬
taining linguistic features of classical Greek, L = dhimotiki), Swiss
German (H = Standard or High German, L = the various dialects of
Swiss German) and Haitian Creole (H = French, L = the local French-
based Creole). According to Ferguson, in diglossic societies speakers
have a clear notion of the appropriate use of each variety. To use the L
in situations where the H is required constitutes, therefore, a serious
social blunder, while the opposite mistake would be an object of ri¬
dicule.
This study was based on the linguistic situation found in countries
where, as pointed out, the two varieties, H and L, belong to the same
language. The notion of diglossia was later revised and expanded by
Fishman (1967), who proposed terms of reference which would allow
the linguist to account for bilingualism and diglossia, and the relation¬
ship between the two, within one conceptual framework. Fishman
maintains that the two phenomena are quite distinct: diglossia is a fea¬
ture of society, to be studied by sociologists (and, presumably,
sociolinguists); bilingualism, in contrast, concerns the individual’s
ability to use different language varieties (or separate languages) and
therefore belongs to the realm of psychological (and psycholinguistic)
investigation. Fishman extends Ferguson’s notion of diglossia by
broadening it to include different dialects, vernaculars or classical var¬
ieties, as well as distinct languages - so long as they are functionally
differentiated. He proposes to accept more than just two varieties
(whereas Ferguson mentioned only the High or super-posed and the
Low or ‘primary dialect’ forms of the language). Fishman represents
the relationship of diglossia in a quadrant diagram, and he then dis¬
cusses each possible construct in some detail.
Diglossia
+ -
(1) Both diglossia (2) Bilingualism

+ and without
bilingualism diglossia
(3) Diglossia (4) Neither

- without diglossia nor
bilingualism bilingualism
Source: (Fishman 1967: 30)
FIGURE 8.1 The relationship between bilingualism and diglossia
In a speech community where both diglossia and bilingualism are

present, almost all speakers will know both the H and the L varieties.
Fishman cites as examples language situations involving Swiss German
and Arabic, and also Guarani (an American Indian language) and Span¬
ish in Paraguay, where Spanish, as the official language, is used in
higher-level education, the media and for formal communication,
whereas Guarani is spoken by nearly all Paraguayans in informal situ-
atiorts. Bilinguahsm without diglossia is found in areas where there are
large numbers of bilinguals who use either language for almost any
purpose. Fishman considers that this situation pertains only to ‘circum¬
stances of rapid social change, of great social unrest, of widespread
abandonment of prior norms before the consolidation of new ones’
(1967. 35). Clearly, when bilingualism is widespread in a community it
is more usual to find social consensus as to the functional and situ¬
ational distribution of languages (i.e. type 1 Figure 8.1). Fasold (1984),
quoting Verdoodt (1972), gives as an example of this type the German-
speaking area of Belgium, where a shift from German to French is
under way but there are still many speakers who use either language
for most purposes.
The third type of relationship, diglossia without bilingualism, exists
only when two or more quite different groups come together within the
same political, economic or religious organization (Fishman uses the
word ‘unity’) but do not have much contact with each other. The ruling
group uses the H variety, to which the other group(s) may have only
restricted access. Fishman points out that there are many examples of
ruling elites who use a different language from those whom they rule -
in economically undeveloped and socially immobilized societies
‘locked into opposite extremes of the social spectrum’ (1967: 34), e.g.
in India, where the caste system contributes to a strict separation of
social groups. In pre-First World War Europe such a linguistic situation
could be found in czarist Russia, where the nobility spoke French and
the rest of the people spoke Russian and other languages. These exam¬
ples, and Fishman’s conception of diglossia without bilingualism, show
that in a type 3 situation there is a fairly rigid separation of the two (or
more) speech communities and in the absence of interaction between
the groups communication is possible only through interpreters or by
using another language altogether, e.g. a lingua franca which may be a
pidgin.
The fourth quadrant in Fishman’s construct tends, in his own words,
to be ‘self-liquidating’, as it is virtually impossible to find speech com¬
munities where only one linguistic variety exists, that is, with no
stylistic variation (which he includes in his concept of diglossia).
Fishman and Ferguson differ in their approach to, and interpretation
of, the notion of diglossia, but they agree on the functional distribution
of language varieties in society. This is the cmcial point if we wish to
make use of the concept in order to classify the various patterns that
are found in the study of societal multilingualism. Fasold (1984) de¬
scribes other forms of diglossia that involve one H and several L
varieties, and he mentions also the possibility of having different
‘layers’ of varieties where the High and the Low forms overlap. His
examples are taken from Tanzania and India, and they emphasize the
role of diglossia in creating bilingual and multilingual language situ¬
ations in those contexts.
8.3.4 Examples of multilingual patterns in contemporary Europe

This section contains an outline of some types of multilingual patterns
found in western Europe. It is important to bear in mind that individual
countries usually include a number of different language varieties (be¬
longing to the same or to distinct languages) within their borders, and
that therefore a multiplicity of language patterns can be found almost
everywhere. On the other hand, language behaviour varies from group
to group, as does the way in which languages are officially recognized
and used for communication and/or education purposes.
170
SOCIETAL MULTILINGUALISM
(1) Switzerland Switzerland (or the Swiss Confederation - of twenty-

three cantons and three demi-cantons) is a small country of some six
million people which has four ‘national’ languages: German, French,
Italian and Rhaeto-Romansch. Only the first three, however, are men¬
tioned in the Swiss Constitution, which means that Romansch does not
enjoy the same rights in the federal parliament and in administration as
the other three. Since Switzerland follows the principle of territorial
monolingualism, the individual cantons are linguistically autonomous.
Most cantons are monolingual, a few are bilingual (e.g. Jura and Frei¬
burg/Fribourg) and the Graubunden canton is trilingual. In the sixteen
German-speaking cantons a range of regional and local dialects (col¬
lectively referred to as Swiss German or Schwyzertiitsch - also spelled
Schwyzerdiitsch) is used; some of them have become widely accepted
beyond their original regions, e.g. Zurichdeutsch (Ris 1979 refers to
these as Grofiraumdialekte). In addition to dialect varieties there is di-
glossia, the Swiss-German dialects being the L form, used for everyday
(mainly spoken) communication, and standard German serving as the
H variety, reserved for formal uses, both oral and written. German or a
German-based dialect can be said to be the first language of more than
two-thirds of Swiss people.
French is dominant in six cantons, three of which are bilingual. Al¬
though High German is taught in schools, few French-speaking Swiss
people learn any Swiss-German dialects, and their representatives in
parliament insist on the use of the H variety (standard German), rather
than Schwyzertutsch, for parliamentary business. French is spoken by
at least 20 per cent of the Swiss population. Italian dialects are spoken
by 10 per cent, principally in the Ticino canton. Standard Italian is
used mainly for wntten purposes. Ticino has experienced a large influx
of non-Itahan (mainly German) speakers who have been reluctant to
adopt the Italian language and have set up their own German-medium
schools, societies and associations.
Rhaeto-Romansch is the mother tongue of only about 1 per cent of
the country’s population, perhaps even less. There are five main dia-
lerts spoken, but no standard language. Owing to Graubunden’s
difficult accessibility, its long history of migration and lack of social
unity, and also because of the recent permeation of the area by German
and Italian speakers, the Grisson (as the region is also called) is today
a trilingual canton where the number of Romansch speakers is steadily
diminishing.
One further point is that Switzerland has the highest proportion of
foreign residents of any country in Europe, perhaps over one million

people, many of whom speak English at least as a business language.
Furthermore, in the last thirty years or so Switzerland has attracted
large numbers of foreign workers who have settled in the country.
(They are, of course, also ‘resident’ but tend to be referred to separate¬
ly, usually as ‘migrant workers’.) Migration has brought yet another
form of language contact, often encompassing both multilingualism
and diglossia.
The language situation in Switzerland is marked by polydialectalism
in all cantons, by monolingualism in most German- and French-speak¬
ing areas, and also by bilingualism and trilingualism in some parts.
Besides all this, there is diglossia in the German-speaking cantons.
(2) Luxembourg The Grand Duchy of Luxembourg is one of Europe’s

smallest nations (population 350,000 approximately) and also one of
the most interesting linguistically. It has been referred to as triglossic,
since three languages figure prominently in the country: German,
French and Letzebuergesch (or Letzeburgisch, or Luxembourgish), an
Alemannic variety related to Alsatian and the Swiss dialects. All three
varieties have remained a stable part of Luxembourg’s language pat¬
tern, and each is used for specific purposes in certain well-defined
situations. German is used as the written norm; for oral communication
it is used only at school, as a language of instruction. Letzebuergesch
is perceived to be a separate language, and it is used by every Luxem-
bourger for virtually all spoken purposes except official speeches in
parliament and in court, where French is used. French is also employed
as a medium of education at secondary level, and it serves as the lan¬
guage of all official written communication. There exists a certain
diglossic relationship between French and Letzebuergesch, and also
with German, but the main feature of Luxembourg’s language situation
is that adults use all three varieties, and that this accords with the
country’s constitutional provision.
(3) Spain The Spanish (1978) Constitution divides the country admin¬
istratively into seventeen ‘autonomous [or self-governing]
communities’, and it allows for those which have their own language
to use it as an ‘official’ language in the respective areas alongside Cas¬
tilian Spanish (the Constitution mentions ‘the languages of Spain’).
While extensive parts of Spain are polydialectal but monolingual, bi¬
lingualism is found in three comunidades autonomas: Catalonia,
Galicia and the Basque Country or Euskadi. In these regions language

choice follows individual rather than societal criteria. For instance,
whether a Basque will use Spanish or Euskera (the Basque language)
depends on aspects such as topic of conversation, interlocutor and situ¬
ation, rather than on commonly agreed social norms. The autonomous
government of Catalonia, the Generalitat, has pursued a linguistic pol¬
icy which aims at furthering the establishment of such norms, as
Catalans tend to put their language on a stronger footing than anyone
else in Spain (see further details in Chapter 13). The significant change
that has taken place since the middle 1970s is that there is now, every¬
where in the country, active encouragement of local languages, dialects
and accents (in education and the media for example), whereas former¬
ly Castilian Spanish was singled out and officially promoted, although
only about one Spaniard in four speaks Spanish with the accent of this
prestige variety.
(4) Denmark The Danish sociolinguist Normann Jprgensen describes

(1984) four kinds of bilingualism found in some industrialized so¬
cieties, such as Denmark and other Scandinavian countries. His first
type relates to the use in the home of a minority language which is
also a ‘world language’, whereas the national language (Danish) is
used outside the family circles; the example he gives is the German¬
speaking minority group living near the southern (Danish-German)
border. The second type of bilingualism found in Denmark concerns
speakers of the national language who have achieved a high degree of
competence in a foreign language (in most cases English) learnt at
school and beyond; he says that many Danes, in addition to their
mother tongue, also master English, which is taught from the fifth
grade and is seen as a highly prestigious language that can be heard on
radio and TV, the cinema, popular music, etc., and which most people
consider to be essential for international business, travel and study pur¬
poses. Jprgensen’s third type involves the use of a classical dialect (of
Danish) as the home language and the national majority language else¬
where; he points out that there are still in Denmark peripheral areas
where a dialect is spoken among members of the community, although
standard Danish is the only variety people employ in schools and for
all official communications. This sort of bilingualism could perhaps
better be termed bidialectalism (Jprgensen does not do so; he takes the
view that the use of two differentiated varieties of the same language is
another form of bilingualism). Jprgensen’s fourth and last type refers to
THE UNSTABLE NATURE OF MULTILINGUAL SITUATIONS 173
the linguistic situation found among migrant workers and refugee

groups in Denmark, whose minority home language is not a world lan¬
guage and who also speak the national language of the host country.
He cites the children of Turkish workers who speak both Turkish and
Danish.
The kind of bilingualism exemplified by Jprgensen’s types 3 and 4
is sometimes referred to as popular bilingualism (see section 8.1.2).
His type 2, which is the result of schooling rather than natural acquisi¬
tion, has been called cultural bilingualism by Skutnabb-Kangas
(1984a).
A fifth type can be added: dite bilingualism. In Europe (and also
elsewhere) we find many bilinguals who are members of international
communities, mainly in metropolitan areas, i.e. people who work for
multinational firms and organizations or world agencies using an inter¬
national language, and whose children are brought up bilingually.
Denmark provides a good example of present-day linguistic patterns
in western Europe. Many countries nowadays have new linguistic mi¬
norities created by migration or immigration, alongside their one or
more old-established linguistic group(s), a situation which results in
different kinds of elite, cultural and popular bilingualism, in addition to
the previously existing bi- or polydialectalism. Education, mobility and
internationalization have all contributed to the emergence of a chang¬
ing linguistic situation, such as seems to be prominent today in some
parts of Scandinavia, the Benelux countries and Switzerland.
8.4 The unstable nature of multilingual situations
A country’s linguistic situation is never static. Changes in its political

and economic fortunes may bring about an alteration of its language
pattern, just as changes in the social conditions under which its inhabi¬
tants live may cause them to modify their linguistic behaviour (e.g. as
a result of migration into or out of a minority area). In multilingual
countries such developments imply that the languages change also in
relation to each other. In the Swiss canton of Ticino, as we saw, tour¬
ism and the arrival of non-Italian speakers have brought about a
situation where the newcomers have not assimilated linguistically to
Italian, the local official language, as was traditionally expected of
those who move from one language area to another. As a result, the
Ticinos are undergoing a remodelling of their linguistic patterns, and
there is some fear that the position of Italian may become further
weakened. Belgium’s capital, Brussels, has undergone a massive lan¬
guage switch in recent times, as a result of urban migration,
centralization of the country’s administration and internationalization.
Whereas 150 years ago the city was predominantly Dutch-speaking (it
is, after all, situated in the Flemish-language region of Belgium), today
its population is mainly French-speaking (Baetens Beardsmore 1983).
Modem western society demands a good deal of conformity from
its citizens. Part of the price that is often exacted by economic progress
seems to be a reduction in linguistic diversity, although there is no
reason to assume an inherent causal link between monolingualism and
prosperity. We can observe that polarization of language use has oc¬
curred virtually everywhere in Europe, as most linguistic minorities
have seen a steady decline of their languages. At the same time, it is
possible to see new linguistic situations arising from changing social
and economic circumstances. Migration and immigration have created
patterns of multilingualism hitherto unknown in most European coun¬
tries, and a revival of regional nationalism in several areas of the old
continent (both east and west) has brought a renewed interest in re¬
gional languages in their wake.
The description and analysis of multilingualism must take account
of the changeable nature of language patterns. Multilingualism and
monolingualism should not be regarded as discrete categories, but
rather as points on a sociolinguistic language continuum. Groupings of
categories along this gradient make up specific speech communities
which, in turn, are composed of individuals who have different patterns
of communicative competence and language behaviour. At both the in¬
dividual and the social levels, bilingualism/multilingualism is a sui
generis phenomenon, subject to its own laws.
Chapter 9
Language choice, language maintenance and

language shift
9.1 Language choice
One way of looking at language use in society is to see it in terms of

making choices. Human communication entails selecting from the lin¬
guistic and stylistic items available, i.e. favouring some and rejecting
others. This choice can be made consciously or unconsciously. Where¬
as monolinguals will choose according to the conventions of the
members of their speech community and their own idiosyncratic pref¬
erences, the speakers of diglossic communities will have additional
choices to make, as they must also decide whether the High or the
Low variety should be employed. Similarly, members of bilingual or
multilingual societies are faced with wider choices in their language
use. In Part I (particularly Chapters 4 and 5) the bilingual individual’s
language use was examined with a view to finding some answers as to
when, why and how Language A, or Language B, or a mixture of both,
is used.
In this chapter the focus is on language choice in bilingual/multilin¬
gual communities. It is obvious that many factors which influence
language choice in the bilingual individual will also affect groups of
such people. But we must distinguish between (1) those bilinguals who
live in multilingual settings, i.e. in places where two or more languages
are used throughout the community, and (2) bilingual persons who are
members of a monolingual larger group or society, that is, people who
have become bilingual as a result of migration, marriage or being the
offspring of couples who use different languages in speaking to them,
but who are not themselves members of a bilingual community (al-
176 LANGUAGE CHOICE, LANGUAGE MAINTENANCE & LANGUAGE SHIFT
though they are members of a bilingual family). The language choice

of these two types of speakers will be determined by different sets of
social, psychological and linguistic factors.
Whether we are considering the bilingual living in a multilingual
society or the bilingual individual in a monolingual setting, their lan¬
guage choice will always presuppose; (1) that more than one language
is available to the speaker; and (2) that (s)he will be sufficiently profi¬
cient in the languages concerned for a genuine choice to present itself.
These prerequisites do not imply equal competence in both languages
on the part of the individual, nor that each member of the bilingual
community is fully bilingual. In the studies that touch upon the subject
of language choice (some of them are referred to later), the members
of language groups were not tested as to their individual proficiency,
because the emphasis was on the group as a whole and its use of the
languages studied. The correlation between knowledge and use of the
language will, however, be relevant in the context of a group’s willing¬
ness to maintain or abandon one of its languages. Clearly, an individual
or a group whose command of a language becomes weak will use it
less often, and this may in turn accelerate language loss. This is true in
general, but we shall see that there are few predictable factors in the
process of language maintenance or loss.
There is available a good amount of research which, explicitly or
implicitly, examines the pattern of language use in bilingual com¬
munities. Much of it focuses also on questions of language
maintenance and shift. In some cases the groups studied are of con¬
siderable size, e.g. Rubin’s (1968; 1970) work on bilingualism in
Paraguay, Fishman’s (1964; 1968) studies of Jewish and Puerto Rican
immigrants to the United States, and Greenfield’s (1972) enquiry into
Puerto Rican bilingualism in New York. In other instances the com¬
munities were smaller, or they were settled in remote or isolated parts
of the world, or a combination of both; for example, Sankoff (1972)
studied the Buang in New Guinea, Gumperz (1978) the Slovenes in
Austria, Gal (1978a and b and 1979) the Hungarians also in Austria,
and Dorian (1981) the Scottish Gaelic community in east Sutherland.
In most of the studies in this latter group the information was gathered
in small rural communities - which are, naturally, more easily ac¬
cessible to researchers working on their own. Sociolinguistic research
in urban areas is more complex, in the sense that many different socio¬
economic and sociocultural variables converge. At least two bilingual
cities have been considered: Baetens Beardsmore (1983) studied Ian-
LANGUAGE CHOICE 177
guage shift in Brussels, and Lieberson (1972) examined language

maintenance in Montreal.
Fasold (1984), in his excellent textbook on sociolinguistics, takes
particular care to explain the methodological approaches adopted by
various researchers into language choice. In his chapter on the topic he
discusses in some detail the contributions of three disciplines, socio¬
logy, social psychology and anthropology, which he illustrates by
reference to studies undertaken in the last twenty-five years. These
three approaches will now be considered in turn, in the hope of bring¬
ing out some of the interesting issues involved in the question of
language choice.
9.1.1 Domain analysis: person, place and topic

The contribution made by sociologists towards an understanding of
multilingualism is reflected both in the research methods that have
been established and in much of the terminology adopted. The starting
point for any study of language choice in multilingual communities is
the recognition that, as this choice does not constitute random deci¬
sions on the part of the speaker, there must be a certain pattern.
A good part of Joshua Fishman’s work in the 1960s was concerned
with analysing and describing patterns of communication within multi¬
lingual groups, and also between them. In one oft-quoted article, ‘Who
speaks what language to whom and when’ (1965a, revised 1972), Fish¬
man says that '’Proper usage dictates that one of the theoretically
co-available languages or varieties will be chosen by particular classes
of interlocutors on particular kinds of occasions to discuss particular
kinds of topics’ (Fishman 1972: 19). The aim of the study was to find
out the descriptive and analytical variables that determine language
choice. Perhaps the most important part of the discussion is the section
concerning the ‘particular kinds of occasions , which Fishman de¬
scribes as ‘domains’ of language behaviour and defines in terms of
‘. . . institutional contexts or socio-ecological co-occurrences. They
attempt to designate the major clusters of interaction situations that
occur in particular multilingual settings’ (1972: 19).
The type and number of domains which have been established by
sociolinguists and used in their research vary somewhat. The designa¬
tion of the relevant domains obviously requires a good deal of inside
knowledge of the communicative behaviour and the sociocultural fea¬
tures of particular speech communities. Fishman mentions the domains
described in an older piece of research, which was carried out by

Schmidt-Rohr (1932) among ‘Auslandsdeutsche’ (i.e. emigre Germans
who, in this particular study, were affected by language shift). The nine
domains outlined by Schmidt-Rohr were; the family; the playground
and the street; the school (language of instruction, subject of instruc¬
tion, language of recreation and entertainment); the church; literature;
the press; the military; the courts; and government administration.
Incidentally, it is curious that Schmidt-Rohr overlooked the work
sphere as a domain. Other studies (e.g. Greenfield 1972 and Parasher
1980, both discussed in Fasold 1984; 183 ff.) limit their enquiries to
fewer domains. Greenfield refers to locations such as home, beach,
church, school, and workplace; and Parasher works with the following
seven domains; family, friendship, neighbourhood, transactions, educa¬
tion, government, employment.
Looking at these lists it should be clear that a ‘domain’ can be seen
as the configuration of at least three component factors;
(1) the participants in a conversation;

(2) the place where it occurs;
(3) the subject under discussion.
The participants or interlocutors will be characterized by such features

as age, sex, social status and socio-economic background, and also by
the kind of relationship existing between them, which will determine
the degree of intimacy or formality that they display in relation to each
other.
There is, of course, a multitude of places or locations where conver¬
sations may take place. Many of them will not have any particular
bearing on bilingual speakers’ language choice, but there are certain
areas where a particular language is more likely to be used than an¬
other, e.g. the home, government offices, etc. It is interesting to see
that an item such as ‘school’, as listed in Schmidt-Rohr’s early study,
already shows that subcategorizations are often necessary.
Topic has long been recognized as an important factor governing
language choice. Individual speakers will usually have preferences for
using a particular language when speaking about a certain subject. This
preference can become particularly clear when the discussion of a topic
leads the speaker(s) to switch from one language to another. There are
many possible reasons for such switches; the speaker may feel more
competent in handling a topic in a particular language, perhaps because
LANGUAGE CHOICE 179
(s)he has learnt the appropriate terminology only in the context of one
language; or (s)he may feel that the other language does not possess
the required terms; or (s)he will somehow consider one language to be
better than another for speaking about a particular subject. It is not the
topic per se which requires the choice of one language rather than
another, but the personal experience and perception the speaker has of
a particular topic. Thus when a student from Wales who was educated
bilingually tells me that ‘you cannot talk about maths or chemistry in
Welsh, but it is all right for music, history or literature’, he is saying
that he has not learnt maths or chemistry through the medium of
Welsh, and that therefore he has no experience of talking about chem¬
istry-related topics in this language, for it is indisputable that Welsh
has the necessary linguistic tools for dealing with these subjects.
So instead of looking individually at topic, place and person, it may
be more interesting to bundle them into situations such as ‘conversing
with friends and acquaintances’, ‘talking to people at social gatherings’
or ‘arguing with friends/colleagues in a heated discussion’ (Parasher
1980). Such situations can then be arranged into the more and the less
formal ones, which may be useful when looking at language choice in
diglossic communities. The typical situations outlined by Ferguson
(1959), where the two varieties. High and Low, are distinguished, are
separated into more formal/less formal ones (see section 8.3.3). In
these situations one can also distinguish the medium employed, i.e.
whether spoken or written language would be used.
Another approach to language choice, similar to domain analysis,
has been adopted by Joan Rubin in her study of bilingualism in Para¬
guay. (Paraguay is a bilingual country; Spanish and Guarani are used
in a diglossic relationship, with Spanish as the High and Guarani as the
Low variety.) She set up a number of categories, indicating both posi¬
tive and negative values. These categories correspond broadly to place,
topic and person.
Rubin found that, broadly speaking, it could be said that a Para¬
guayan’s decision to use Guarani or Spanish was based on a series of
considerations which could be ordered in a specific hierarchical way.
This kind of representation of ordered language choice is called a ‘de¬
cision tree’. It shows the sequence of choices a Paraguayan bilingual
will go through when deciding whether to use Guarani or Spanish.
First (s)he will consider whether or not (s)he is in a rural location: if
(s)he is. Guarani is the likely choice; if (s)he is not, then the formality
of the situation may decide; if it is formal, Spanish will be appropriate;
Location
Rural Guarani Non-rural
1
Formality-informality
Formal Spanish Non-formal

i
Intimate
Non-intimate Intimate
Spanish
I
Seriousness of
Non-serious Serious
Guarani
First language learned

Predicted language
proficiency
Source: Joan Rubin, 1970; in Fishman, Readings in the Sociology of
Language, 1970: 526
FIGURE 9.1 National bilingualism in Paraguay; ordered dimensions in the

choice of language in a diglossic society
if it is not, the additional consideration of a relationship between the

speakers may bring about a decision - if there is no intimacy, Spanish
is the likely language to be chosen, but if there is a degree of intimacy
other factors must be taken into account. Such a decision tree, if based
on accurate observation and statistical data, can then be used also to
predict language choice, not only to explain it. However, not all deci¬
sions on language use can be attributed to factors determined by social
stmcture.
9.1.2 The socio-psychological approach: linguistic accommodation

The social psychologist approaches the study of human communication
from a different angle. (S)he claims that not only social considerations.
LANGUAGE CHOICE 181
but also factors relating to the psychological forces which influence the
individual’s actions, need to be taken into account. For instance, the
individual’s desire to identify with, or dissociate from, a particular lan¬
guage group can be a determining factor in language choice. Howard
Giles and his associates (Giles, Bourhis and Taylor 1977) looked at
language choice in terms of the individual’s desire to emphasize or
weaken her/his ties with the respective language groups. The basic idea
of their accommodation theory is that normally a speaker will ‘con¬
verge’ and choose the language which suits the needs of the
interlocutor. The opposite decision, ‘divergence’, would represent a
conscious decision not to adjust one’s speech to the person one is talk¬
ing to. Linguistic adjustment can take various forms. It may be a total
switch from one language to another (one of which the interlocutor has
a better command), or it may involve speaking more slowly, pronounc¬
ing the words more emphatically; speakers may also choose to use
more elements such as pronunciation features, or words and express¬
ions from the other language, or short-passage translations. The main
emphasis is on the speakers/listeners as members of groups, and the
terms used for the languages involved are ‘ingroup language’ and ‘out¬
group language’. The factors that influence convergence or divergence
depend on whether or not:
(1) the interlocutors see themselves as members of the same group;

(2) they would like to be considered as members of the other’s
group;
(3) conflict exists between the groups;
(4) the two groups are equally aware of a conflict, if it exists.
Multilingual communities more often than not consist of one domi¬

nant group and one or more subordinate groups, the former being
characterized by being in possession of the economic and political
power of the country and (particularly in highly centralized states)
being in a position to impo.se their language upon the subgroup. The
subgroup’s language behaviour will typically include linguistic accom¬
modation (convergence) in ingroup-outgroup contact situations. The
pattern of convergence will usually involve movement in one direction
only, namely speakers of the subgroup converging towards the lan¬
guage of the dominant group; e.g. members of the Italian community
in Britain will use their ‘best’ English when communicating with their
Bank managers (or similar formal situations). However, any attempt on
the latter’s part to converge by, for example, adjusting their pronunci¬
ation to Italian or using Italian words would probably be construed as
mockery.
Linguistic accommodation by subordinate groups is very common,
but it is not an inevitable occurrence. Some linguistic minorities in
Europe are gaining more self-confidence, which is coupled in some
instances with the attainment of greater recognition or even a degree of
self-government. It has been observed (for example in Wales, South
Tyrol and Catalonia) that speakers of the minority language show less
willingness to switch automatically to the majority language when
speaking to a member of that speech community. Thus, a change in
accommodation patterns may indicate that a social change is under
way. When more and more Catalans use their language in situations
where Castilian Spanish was formerly the norm (e.g. in public adminis¬
tration), and when they begin to demand convergence from the former
dominant group (now that Catalonia has achieved autonomy within the
Spanish state, the aim is to make Catalan the language of the school,
which means that the children of non-Catalan homes have to learn it),
then this is evidence of a social change; that is to say, the former status
as subordinate group changes to that of equal-status group.
9.1.3 Anthropological methods of observing language choice

The'anthropologist’s contribution towards the study of language choice
can be seen in terms of orientation and methodology. Language choice,
in the anthropologist’s opinion, should reveal something about the atti¬
tudes and the cultural values held by speakers in multilingual
communities. Linguistic behaviour and the demonstration of cultural
values (explicitly or implicitly) can be ascertained only by observation.
Therefore, sociologists and psychologists use data collected from cen¬
suses, interviews and questionnaires, or look at people’s behaviour
under controlled experiments. The information obtained is then usually
presented in numerical form and statistically analysed, in the hope that
the results will indicate significant trends. The anthropologist, on the
other hand, needs to observe her/his informants in as many normally
occurring situations as possible, as the aim is to gain insights from
uncontrolled behaviour. The methodology that has been adopted in
many cases is called participant observation, which requires the re¬
searcher to live among the members of the community under
investigation and to take part in their daily lives. Joan Rubin (1968),
LANGUAGE CHOICE 183
Susan Gal (1978a and b; 1979) and Gillian Sankoff (1980) all spent
long periods of time living among the speakers whose language beha¬
viour they studied, and Nancy Dorian (1981) spent over a decade
working on language shift in northern Scotland. It is understandable
that research which combines data collected through interviews, ques¬
tionnaires and records of actual observation should prove particularly
attractive. If the results achieved by each method are kept separate
until the end of the study, then compared with each other and found to
be consistent and mutually corroborating, such research can be said to
have high validity. But statistical analysis is not usually employed in
this kind of anthropological research, and the main emphasis remains
on the observers’ intensive involvement with the communities they are
studying; any statistical data that are collected may be used only as
supplementary material.
Fasold (1984: 193) makes the point that anthropological work on
language choice has thrown considerable light on community structure.
He develops a model of basically two types of structure that can be
found in certain bilingual communities.
The first type would be one in which the two communities (one the
ruling elite and the other the governed group) exist side by side with¬
out much contact, except the minimum necessary, perhaps by means of
a pidgin. The members of each group do not need to be bilingual, and
both groups perceive themselves as separate from each other. This kind
of structure is not frequently found in the world, although some so¬
cieties do approach this model to some degree; but it is useful to bear
it in mind for contrasting purposes.
The second type of community structure can be seen, for instance,
in a minority language community such as the Frisians in the Nether¬
lands or the Slovenes in Austria. This structure represents a social
arrangement that is different from the previous one. As in the first case,
one group is the dominant one, with more power and prestige, using
the majority language, and the other is the low-status one, which habit¬
ually uses the minority language. However, this group does see itself
as part of the other, i.e. as included in the high-status group. Member¬
ship of this ‘dual group’ can have important linguistic consequences.
For example, members of the low-status group will tend to be mono¬
lingual in the minority variety at the beginning of their life; but, as
they grow older, their contact with the majority variety may increase
(e.g. through formal education), and they may become bilingual. At the
same time they may develop allegiances towards both their original
group and the other social group, which is perceived as including their
own. The language choices of the members of these bilingual groups
may show the kind of allegiance the speakers feel towards each group.
But the dual loyalties can also cause some degree of conflict. Attempts
at solving the problems arising can take various forms and may have
interesting linguistic consequences.
An instance of such a linguistic strategy, attempting to resolve con¬
flict between two varieties, is the gradual change from one variety to
another within the same interaction (see an example in Blom and
Gumperz 1972). Another example could be the sudden switch from the
minority language to the majority upon demand from, or in the
presence of, monolingual members of the majority variety (as reported
by Gal 1979: 166). Gal’s work shows several types of language
choices which reflect the group dynamics of the community she was
studying in Oberwart, Austria, near the Hungarian border, where a
Hungarian-speaking population has existed for many years. Until the
last century this area was mral and isolated, and the Hungarian spea¬
kers had little contact with the German-speaking Austrians. Political
and social changes which have taken place in the present century have
led to a situation where German speakers have become the majority
and, because of the high status of German as the official language and
the medium of education in Austria, Hungarian has come to be re¬
garded as the low prestige variety, spoken by the members of the
community who see themselves as Hungarian Austrians and who are
bilingual. Gal found that age was an important indicator of whether or
not a member of the Oberwart community would speak Hungarian.
The general pattern was that the older members of the bilingual group
spoke more Hungarian than the younger ones, and also that the latter
spoke German more fluently than their elders. An even more important
finding was that language choice seemed determined by the interlocu¬
tor, rather than by any other factor. Even the younger members whose
parents did not speak Hungarian to them reported that they used this
language in hymns and prayers addressed to God (Gal did classify God
as an interlocutor); many also used Hungarian to their grandparents
and members of the older generation. The use of Hungarian for relig¬
ious purposes can be explained in terms of situation as well, because in
Oberwart almost all the members of the Hungarian-speaking com¬
munity are Calvinists, and Hungarian is the language in their churches.
The use of Hungarian in other situations, on the other hand, depended,
according to Gal’s observations, on the participants involved and, to
LANGUAGE MAINTENANCE 185
some extent, on whether monolingual members of the community were

present.
We can therefore say that language choice is determined by linguis¬
tic considerations as well as by external and internal factors. Among
the linguistic determinants we can include the codes available to a bi¬
lingual community and the degree of fluency the speakers have in each
of them. Location (place and setting are other terms used), community
structure, situation (or degree of formality), topic and participants (or
interlocutors) are the variables which have been most frequently dis¬
cussed by researchers, who often approach the subject of language
choice from different angles (corresponding to their academic lean¬
ings), emphasize diverse aspects and use distinct methodologies.
9.2 Language maintenance
We have seen how the observation and analysis of linguistic choices

within a multilingual community can provide us with insights into the
social structure of the group and into the dynamic forces that regulate
their speech behaviour at a given time. We now come to consider the
study of language maintenance and language shift, which can, in addi¬
tion, point to social changes that have taken place, over a period of
time, in the community under observation. Language maintenance and
language shift should not be seen as complete opposites: when one
focuses on one the other may still be present, albeit less prominently.
The expression language maintenance refers to a situation where
members of a community try to keep the language(s) they have always
used, i.e. to retain the same patterns of language choice. Language
maintenance can thus be said to reflect collective volition. In a multi¬
lingual community this may find expression in a group s conscious
efforts to protect its language and ensure its continued use. Such
measures reflect the group’s self-consciousness and also some degree
of political independence, which allows it to determine its own lan¬
guage policies. When language maintenance efforts result in the use of
the minority language in the realms of administration, education and
industry and business (on the basis of a largely monolingual infrastruc¬
ture), then such efforts can be maximally effective. This type of policy
(which Nelde, 1984, calls the ‘Belgian cure’) has been practised in
French-speaking Canada and in Catalonia with a good measure of suc¬
cess. In multilingual countries language maintenance over long periods
of time can be a strong indicator that each of the groups involved con¬
siders its language to be an important feature of its ethnic or national
identity. Unless the groups are very isolated from each other, it will
usually be a sign, also, of a stable social arrangement whereby the
component groups enjoy equal status or some degree of self-determina¬
tion. Switzerland is a good example of such a multilingual country,
since each of the four language groups has maintained its language
over a long period of time. However, close inspection of the two smal¬
ler groups, the Italian- and the Romansch-speaking communities,
shows that some degree of language shift has taken place in spite of
the collective decision of the respective peoples concerned to retain
their language.
9.3 Language shift and its causes
When a community does not maintain its language, but gradually

adopts another one, we talk about language shift. The shift can be
complete, involving a change from one kind of monolingualism to an¬
other, with only the transitional period marked by group bilingualism.
The change from Cornish to English in Cornwall can serve as an illus¬
tration. The shift can also be incomplete, as for instance in those cases
where a section of a speech community retains the old language in
additidn to acquiring the new one; this is the case in Wales, where
everybody speaks English but about one-fifth of the population can
also speak Welsh. Another form of incomplete shift can be that all (or
a large majority of) the members of the community maintain some
degree of proficiency in the language, because they continue to use it
for certain functions; we saw in the previous section the example of
the bilingual community in Oberwart, where Hungarian continues to be
used as the language of religious worship.
The distinction between language shift and language change is one
of viewpoint rather than substance. The latter is a phenomenon of par¬
ticular interest to linguists who look at the small phonetic, lexical,
syntactic and stylistic changes continuously taking place within one
language system. The study of language shift in a community always
entails examining language use in two (or more) languages. The sub¬
ject offers two broad areas of interest, the linguistic and the
sociohnguistic. The codes involved in language shift often experience
changes, such as loss of inflections, borrowing, and appearance of new
LANGUAGE SHIFT AND ITS CAUSES 187
pronunciation features, which can be the object of the attention of the

linguist. The sociolinguist is more interested in finding out why and
how communities undergoing language shift change the pattern of their
language use.
Language shift is sometimes also called language decline and lan¬
guage death. Both terms seem to imply a considerable degree of
personalization. Typically, they would be applied to situations where
the language is spoken by fewer and fewer people until it is no longer
spoken by any member of that community, nor by any other group
outside it; the language disappears with its last speaker. Perhaps the
terms imply also that those who use them regret the passing of the
language(s) in question into oblivion, as the use of the word ‘death’
suggests that there was a previous state of ‘living’. But languages in
themselves are not living things, and they are not ‘brought to life’ by
their users, either. They are inherited from previous speakers. Edwards
(1985; 49) expresses this idea succinctly when he says; ‘Languages do
not possess “an inner principle of life”, nor do they have intrinsic
qualities which bear upon any sort of linguistic survival of the fittest.’
Languages are tools for human communication, and under certain cul¬
tural, social and political conditions a community may opt to change
one set of linguistic tools for another. The change may be accompa¬
nied, or accelerated, by sheer external pressure, but it may also be the
effect of more subtle causes.
Language shift has always been a common occurrence all over the
world. In Africa, America and Asia the process was speeded up by
colonization and, more recently, state-building; both created favourable
conditions for the spread of the ‘big’ colonial languages (e.g. English
or Spanish). Many American Indian languages disappeared within a
period of 200-250 years without any trace being left as no written
records of note existed.
In western Europe relatively few languages have become totally ex¬
tinct, but there are quite a few ‘endangered species’, i.e. languages that
are in the process of being abandoned in favour of another. Examples
are Frisian, Romansch and the Celtic languages (Scottish Gaelic, Irish
Gaelic, Welsh and Breton - Manx and Cornish no longer have any
native speakers). Many studies of linguistic minorities consider ques¬
tions such as whether a minority language is ‘at risk’, that is, whether
it is being maintained adequately (e.g. in Belgium) alongside the lan¬
guage of the majority and whether efforts are being made to recover
(as in Catalonia) or even revive (as in the Irish Republic) the language.
188 LANGUAGE CHOICE, LANGUAGE MAINTENANCE & LANGUAGE SHIET
Many cases of language shift that have been studied involve com¬
munities that constitute a minority within the state and consider
themselves part of that state, but whose language is seen as the low-
prestige variety. In the European context this generally means that the
language of the majority is also the national language. However, there
are exceptions; for instance, in Finland both Finnish and Swedish
(which is spoken by fewer people there than Finnish) are national lan¬
guages; and in some countries the minority language has official status
in the minority area, as German in east Belgium, Catalan in Catalonia
and Welsh in Wales.
Bilingualism does not inevitably lead to shift. There are many bi¬
lingual/multilingual communities that have had perfectly stable
language situations for a long time. In fact, this century has seen a
considerable increase in bilingualism and multilingualism, as more and
more countries have adopted English, and to a lesser extent French
(e.g. in Cameroon) and Russian (e.g. in the USSR) as a second lan¬
guage. At the moment it looks quite unlikely that such countries (in
eastern Europe and Asia, and particularly in Africa) will ever want to
give up either their first or the new second language.
The process of changing from one language to another does not
often occur within the life span of one person; it is usually a matter of
several generations. But it can happen within one’s lifetime, as in the
occasional cases of abrupt shift caused by military conquest leading to
extermination of a group along with its language. One does also come
across some cases of people who, in the course of their lives, have
given up the language of their childhood, never used it later in life and
eventually lost the ability to speak it. Among immigrant groups who
were anxious to assimilate quickly to their host country and/or who
(for whatever reason) took the decision to speak the ‘new’ language to
their children, the languages from their countries of origin survived in
some form for at least two generations. Since it is a gradual process,
language shift has to be studied over a long period of time. Data from
the past, such as censuses or other official records, need to be con¬
sulted, and observation (if that method is adopted) has to span years
rather than months. Only with the appropriate perspective is it possible
to see unequivocal signs that change is under way; normally, the first
symptoms are that the new language acquires functions and uses that
were formerly the attributes of the old.
But why do some communities become involved in language shift,
whereas others maintain stable bilingualism? There is general agree-
ment in the literature on the subject as to the conditions that may bring
about language shift. Not all of them may cause change in all com¬
munities, as the prevailing circumstances are never identical. This is
also the reason why one can only speculate about the imminent onset
of shift, but never reliably predict it. The most frequently mentioned
causes of language shift are the following.
9.3.1 Migration
The members of language groups leave the area where their language
is spoken by the majority of the population (or at least by a sizeable
minority). Usually they move to a part of the world where their lan¬
guage does not serve them any longer, and they adopt the language of
the new area. European immigrants (from both majority and minority
language groups) to the United States are a valid example. Conversely,
the area may attract large numbers of migrants, who bring their lan¬
guage with them and spread it among the local population. This
happened, for instance, when many English speakers moved into south
Wales in the wake of the Industrial Revolution. In Catalonia immigra¬
tion is considered the biggest stumbling block in the present efforts to
recover Catalan, as the influx of non-Catalan speakers into the area
earlier this century is proving a considerable obstacle in the attempt. A
mixture of in- and out-migration has contributed to weaken the Ro-
mansch-speaking community. Many Romansch speakers have to
migrate in search of employment, while at the same time the tourist
industry is being developed in their homeland (the canton of Graubiin-
den in Switzerland) by entrepreneurs who bring their capital and their
language from the German-speaking part of the country. During the
tourist season large numbers of non-Romansch speakers fill the resorts
of GraubUnden, requiring services from the local bilingual population
in German or Italian, not in Romansch.
9.3.2 Industrialization
Industrialization has always triggered off migratory movements, and
this can adversely affect the linguistic stability of an area, in the same
way as other major economic changes. Nancy Dorian’s study (1981) of
the disappearance of the Gaelic dialects spoken in east Sutherland, in
Scotland, shows the relationship between the decline of the local fish¬
ing industry and the language of the fishermen (Gaelic). Similarly,
industrialization brought large numbers of speakers of Castilian Span-
ish to the Basque Country (Carr 1973), just as it brought French spea¬
kers to Brittany (Timm 1980). The vulnerability of minorities to
sudden economic changes was demonstrated again recently in Scandi¬
navia. The Same (often called Lapps by outsiders) have traditionally
been a nomadic people who made a living from breeding reindeer
which they moved around various parts of northern Sweden, Finland
and Norway. After the Chernobyl disaster in May 1986, the grazing
grounds of their herds became seriously contaminated by radioactivity
and are likely to remain unsuitable as pastures for at least a generation.
Thus the whole life style of these people was affected in such a way
that many had to give up reindeer herding, disperse through other parts
of the nordic countries and settle among speakers of Finnish, Swedish
and Norwegian. It is at the very least questionable whether the Same
language will survive as the living idiom of a community much be¬
yond the year 2000.
9,3.3 Urbanization
Urbanization is a phenomenon related to both migration and industrial¬
ization. The movement of sections of the rural population into towns
and cities, linked with the improvement in transport and communica¬
tion systems, contributes to the dispersal of linguistic communities and
brings them into increased contact with the high-prestige language or
otherNinguistic groups. The lack of a linguistic heartland greatly wea¬
kens the survival of a low-status language in a bilingual community, as
life in modem urban societies favours monolingualism, i.e. the use of
the high-status language only. The effects of industrialization and ur¬
banization upon the use of Alsatian in cities such as Strasbourg and of
Catalan m the Barcelona conurbation are discussed in Chapters 12 and
9.3.4 Prestige
Another factor which often (but not always) comes into play is the
small size (m relation to the larger group) of a speech community that
is in the process of changing its language for that of the majority, with
the consequent higher prestige of the new language. Studies by Gal
1979 (of Hungarian in Austria), Dorian 1981 (of Gaelic in Scotland),
Gumperz 1977 (of Slovene in Austria), Walker 1987 (of Frisian in
Schleswig-Holstein, Germany), and Greene 1981 (of Faroese and Cel¬
tic languages) all describe instances of language shifts concerning
small, low-status groups who are in the process of adopting the high-
prestige variety. In many bilingual communities personal wealth,
professional standing and general technological advance are seen as
attributes of the high-status group, so that if members of the minority
language group wish for upward mobility, improved living standards
and/or a share of power, they relate these feelings to the need - which
may be real or simply perceived — to change one’s language. For the
use of the native language makes people instantly recognizable as
members of communities which may be regarded as rural and back¬
ward, suspicious of modernization and political power.
9.3.5 Use as school language

One of the most powerful causes of language shift can be seen in those
areas where the school language is that of the high-status group and no
provision is made for the children of the low-status group to learn to
read and write the language of their ancestors. Unless they receive
good-quality bilingual education, the survival of their first language
will be in question. Monolingual schools (where only the low-status
language is used) have little attraction for a population that is well
aware of the need to be able to master the language of the dominant
group.
As was pointed out earlier, language shift can be brought about as a
result of pressure being applied to a community by the dominant
group, but this is not the only way in which a shift may take place.
Sociolinguists have pondered the question of whether languages are
‘murdered’ or ‘commit suicide’ (for a discussion with reference to lan¬
guage shift in the Irish Republic, see Edwards 1985). European history,
is full of examples of state suppression of its linguistic minorities. In
actual fact, there is hardly a European state not guilty of having in¬
dulged in attempts of this kind at some point in its past. But the
survival of so many languages in spite of prolonged suppression shows
that it may not be possible to eradicate a language if its speakers are
determined to keep it alive, and if they see its continued use as an
effective means of defying the oppressor. Nationalism and separatism
may resist ‘language murder’ attempts in some cases. But in many
others it is not so much a question of open hostility on the part of the
dominant group, but rather one of ignorance and neglect on every¬
body’s part. People who describe themselves as liberal-minded
members of the majority may express their tolerance for the use of the
minority language, even their admiration of it. But they frequently see
no need to go any further. The survival of a minority language depends
on a great deal more than kind words. Similarly, ignorance on the part
of the minority group can speed up language shift. Admiration for the
dominant group and the adoption of their values, mixed with a degree
of pragmatism, can be equated by the cynic with ‘language suicide’.
Minority members may feel a strong attachment to the language of
their forefathers and still not pass it on to their children. If the senti¬
ment is not matched by actual language use, the consequences are
clear, once a language ceases to be the language spoken at home, its
continued existence will be seriously threatened.
Chapter 10
Language and national indentity
The topics to be examined in this chapter relate to multilingualism in

that they touch on issues of central importance to the language groups
that make up multilingual societies. The relationship between language
and group identity has long been deemed to be of considerable signi¬
ficance in the context of the emergence of modem states, both in
Europe and in post-colonial America, Africa and Asia. When language
comes to be considered an essential component of national or regional
identity, either by individual language groups or by a state comprising
two or more such groups, then certain consequences flow from this
situation - for instance, language planning policies have to be formu¬
lated, accepted and carried out so that national harmony can be
maintained and the smooth running of the affairs of the state ensured.
The study of the relationship between language and nationalism has,
however, proved to be difficult, as the issues involved, particularly
those relating to ethnic or national identity, are complex, and, quite
often, intangible. On the other hand, the subject of language planning
is more accessible, since planning strategies normally become evident
in government action.
10.1 Language and nationalism
Edwards (1985: 1-2) mentions that the relationship between language

and identity has been touched upon within a number of disciplines
(history, sociology, psychology, linguistics, education and others).
While acknowledging that the various perspectives help to see the find¬
ings presented in different lights, he criticizes the tact that many
researchers have not taken into consideration the work done in neigh-
194 LANGUAGE AND NATIONAL IDENTITY
bouring fields. Thus the study of political, economic and social history
has rarely included the treatment of linguistic issues. Perhaps the rea¬
son for this was that, for many years, language was thought to belong
to the realm of philology, where it was largely viewed in terms of the
linguistic changes that have taken place over the centuries. Another
difficulty that Edwards highlights is the nature of the subject. He
makes the point that to talk about identity often raises emotional is¬
sues; while there is nothing wrong with a polemical treatment of the
subject, he argues that this is sometimes done under the guise of objec¬
tive social analysis.
There exist a number of markers of group identity, such as age, sex,
social group, religion, geography, cultural traditions and race. Lan¬
guage is only one of them. Many people (particularly if they are
interested parties, e.g. linguists, or speakers of the languages con¬
cerned) claim that language is the most important marker, for they see
the continued use of their linguistic code as an essential condition for
maintaining their group identity. As will become apparent later, this
may be an extreme view. But language may be a significant indicator
of the vitality of a group where it is part of the core system of values
of that community.
10.1.1 Basic concepts

In what follows ethnicity and ethnic identity are treated as essentially
the same construct. Ethnic group is used to refer to a group of people
who share a feeling of ethnic identity. The three notions have been the
object of attention in many studies, often in relation to nationalism, and
sometimes in connection with minorities. But definitions are seldom
offered. Isajiw (1980) examined sixty-five studies of ethnicity and
found that only thirteen of them gave an explicit definition. As will be
seen from the attempts at defining the concepts, there is something
elusive about the phenomenon of ethnicity or ethnic identity - and, by
extension, also about ethnic group - which explains the reluctance of
some writers to commit themselves to a statement of meaning.
Edwards (1985) mentions some definitions which see nationalism as
an extension of ethnicity, e.g. Francis (1947; 1976), who refers to eth¬
nicity as nationalism which is not completely self-aware, and Weber
a 968), who notes that the presence of ethnic solidarity does not in
itself constitute a nation. In other words, ethnicity is seen as an ingre¬
dient of nationalism, a prerequisite or a preliminary stage. Likewise,
LANGUAGE AND NATIONALISM 195
several ethnic groups (as in many countries in Africa) can make up

what we call a ‘nation’; this is the nation tied to the state (see later). In
an earlier treatment of the subject, Edwards (1977: 254) suggested that
‘at a very simple level, ethnicity can be thought of as a sense of group
identity deriving from real or perceived common bonds such as lan¬
guage, race or religion.’ He concedes, however, that such a general
definition invites more questions than it answers - questions relating,
for instance, to the nature and relative significance of these common
bonds. The definition that he puts forward in his later treatment of the
subject (quoted below) is considerably longer, and it takes into account
many of the factors which emerge in the majority of discussions on the
subject. The first point to note is that ethnicity should not be taken to
imply minority group identity or the identity of a social sub-group. In
the immigrant context, particularly the North American one, this equa¬
tion has often been made, perhaps because a majority group does not
usually define itself in terms of ethnicity, but rather by reference to
some aspect of its superior status (often its political power vis-d-vis
another group). But there is no intrinsic link between ethnicity and
minority.
It is also important to emphasize that there are two different sides to
the subject of ethnic groups: (1) the particular features that they may
show; and (2) the boundaries that set them off from other ethnic
groups. In considering these two aspects, we will find that, even
though cultural or other ethnic features may change over generations,
the ethnic groups tend to remain intact as long as the perceived group
boundaries stay. A group can, for instance, change its language, i.e.
lose its original language and, for whatever reason, adopt another one.
But the lost language may still retain symbolic value for that group and
therefore contribute to the maintenance of boundaries. A case in point
here is the Republic of Ireland where, over the centuries, English has
become more and more the language of daily communication, but Irish
Gaelic, although used in very few situations (mainly for ceremonial
functions), is retained as the country’s national language because of its
symbolic value.
The next factor that must be considered relates to the necessity of
incorporating both objective and subjective approaches into the discus¬
sion of ethnic identity. The objective approach looks at certain
observable features (linguistic, religious, racial, etc.) and defines eth¬
nicity in terms of a ‘given’ historical fact, as something into which
group members are bom. Edwards (1985: 8) argues that it is useful to
do this ‘as a quick means of categorisation’. But it falls short of ex¬

plaining the persistence of ethnicity across generations. In order to
answer the question as to the nature of the basis on which ethnicity is
maintained, a subjective viewpoint should be adopted. Such an ap¬
proach recognizes the significance of perceived features (as different
from objectively observed ones). A number of definitions reflecting
this kind of perspective stress that ethnicity is a matter of belief in
common descent and presumed identity (e.g. Shibutani and Kwan
1965: 40-1; Weber 1968: 389; Fishman 1972b). We can readily accept
that all feelings of allegiance and loyalty are highly subjective.
Here, then, is the definition of ethnic identity put forward by Ed¬
wards (1985: 10):
Ethnic identity is allegiance to a group - large or small, socially dominant

or subordinate - with which one has ancestral links. There is no necessity
for a continuation, over generations, of the same socialisation or cultural
patterns, but some sense of a group boundary must persist. This can be sus¬
tained by shared objective characteristics (language, religion etc.) or by
more subjective contributions to the sense of ‘groupness’, or by some com¬
bination of both. Symbolic or subjective attachments must relate, at
however distant a remove, to an observably real past.
As ethnic groups we can consider those which are bound together

by feelings of shared ethnic identity. Smith (1982: 147) proposes a
definition of an ethnic community as:
... a group of people who possess a myth of common ancestry, a shared

history, one or more elements of common culture and a sense of solidarity.
In all parts of the world we find both ethnic groups whose geo¬
graphical origins lie within the country in which they reside, i.e.
indigenous ethnic groups, and those who originated from a region
which lies beyond the frontiers of their present abode. In the British
context, examples of the former would be the English, Gaels and
Welsh, and of the latter the Chinese, Afro-Caribbeans and Asians who
now live in the British Isles. Many other examples could be given from
other parts of Europe.
The close relationship between ethnicity and nationalism has al¬
ready been mentioned. Ethnic groups are seen as preliminary stages of
nationalities, i.e. groups which are ‘simpler, smaller, more particularis¬
tic, more localistic’ (Fishman 1972b: 3). Nationalism is.
correspondingly, an expanded version of ethnicity. The similarities be¬

tween the two are of substance, whereas the main difference is one of
scale. In ordinary usage, a nationality is a body of people who share
common features like language, history and cultural tradition. Fishman
sees nationalities as ‘sociocultural units that have developed beyond
primarily local self-concepts, concerns and integrative bonds’ (Fishman
1972b: 3).
In his consideration, the concept of nationality does not imply the
possession of autonomous territory. It is true, however, that many
others who discuss nationalism as the basic notion that sustains nation¬
alities include the idea that nationalities are linked to a particular
territory and that they have a claim to, or at least a desire for, self-
determination or some form of autonomous government. For instance,
Kedourie (1961: 9) includes the feeling that ‘the only legitimate type
of government is national self-government’ in his description of nation¬
alism, and Edwards (1985: 13) sees the existence of ‘an idea, the
conscious wish for self-control’ as the essential difference between eth¬
nic and national groups. The characteristic features of nationalism,
then, must comprise a sense of community, and they may also include
the desire for at least some degree of self-determination and the wish
to strengthen and preserve the group ties in relation to other groups.
Objective features such as religion or language are considered second¬
ary to the sense of a group’s own perceived self-consciousness and
uniqueness.
The term state is slightly more specific than ‘country’, as it refers
to a political and territorial unit which is defined on the basis of its
geographical frontiers and administrative autonomy. The use of the
word ‘state’ implies nothing about the ethnic composition of the unit.
It may be ethnically homogeneous (as is, for instance, the Irish Re¬
public) or ethnically diverse (as is Yugoslavia).
TTie meaning of nation is less concrete, in that both subjective and
objective considerations come into play. A nation can be characterized
with respect to the psychological ties and the nationalistic sentiments
that keep its members together. It can, however, be difficult to identify
its component features: we would be hard-pressed to describe the dif¬
ference between, say, the Icelandic and the Portuguese nations, unless
we resorted to looking at their individual ethnic traits, in so far as they
are discernible.
Care should be taken not to confuse ‘state’ and ‘nation’. The term
‘state’ is used in political science to refer to a politically organized
body which is defined with respect to its sovereign status, its territory
and its people. A state may contain more than one nationality (which,
in turn, may consist of several ethnic groups). The concept of ‘nation-
state’ (a political unit comprising one homogeneous national group)
refers to an ideal that was promoted by some philosophers from the
seventeeth century on. In the nineteenth century it was used as an ex¬
cuse for the suppression of smaller nations within a state. But a nation¬
state is seldom a naturally occurring phenomenon, as many states con¬
tain within their boundaries more than one nation. Thus the expression
‘United Nations’ is really a misnomer, as it is an organization that in¬
cludes, as its members, states, not nations. On the other hand, the body
representing those European nationalities which do not enjoy some de¬
gree of autonomy bears its name, ‘Bureau for Unrepresented Nations’,
correctly.
Keeping the distinction between state and nation in mind, and turn¬
ing back for a moment to the subject of nationalism, it should now be
clear that it is nations, not states, that are usually the object of nationa¬
list sentiments. (However, nationalists also typically aspire to having a
state through which to exercise self-determination.) Only in nation¬
states would the citizens’ loyalties towards the nation coincide with
allegiance to the state. It follows, then, that (for instance) a Welsh per¬
son’s feelings of loyalty towards the British state are not the same as
herôr his nationalistic feelings towards Wales; indeed, that person’s
feelings may, in some way, be in conflict. Another example is Catalo¬
nia, where nowadays many Catalans claim to be Catalan rather than
declare themselves Spanish (which often causes annoyance to other
Spaniards). The fact that the feelings towards the state and the nation
may be conflicting is one of the reasons for the frequent tension be¬
tween the two; and it can help explain (but not justify) the eagerness of
states to suppress the expression of nationalist feelings, including lan¬
guage.
It is important to be aware of the way in which the term nationalism is
used in sociolinguistics, because in its general use in other contexts it has,
in modem times, acquired negative connotations precisely because of its
association with states rather than nations and with political right-wing
extremism. This became particularly clear in the case of facist Germany,
Italy and Spain.
The main difference, therefore, between an ethnic group and a na¬
tion is that nations are bigger, more self-aware groups who share a
desire to determine their own political destiny; this may not be the case
with ethnic groups. It is precisely because a political element is in¬

volved that multinational states tend to be politically less stable than
multiethnic nations. It is difficult to say when exactly an ethnic group
ceases to be one and starts becoming a nationality. In sociolinguistics,
contrasting concepts often allow themselves to be seen as points on a
continuum, rather than discrete entities. Norway and Austria are exam¬
ples of states clearly consisting of one nationality each but containing
several ethnic groups. States such as Spain, France and Britain are
more difficult to classify, as they are comprised of more than one na¬
tionality (the Welsh, the Catalans, the Bretons, possibly the Scots,
claim to constitute separate nationalities) and also various ethnic
groups (e.g. Moroccans, Pakistanis, Latin Americans - with their sub¬
divisions - living in these countries, etc.). India and the USSR could
serve as examples of multinational states that also contain different
ethnic groups.
10.1.2 Language and nationalism

In facing the question of whether one is dealing with a nationality or
an ethnic group, language can be a useful indicator. In themselves, the
two concepts do not require language as a component. But the extent
to which a group is prepared to defend and maintain the use of its
language, or the ease with which it will acquiesce to abandon it, may
be significant.
The connection between language and nationalism has often been
noted. Language is a tangible, immediately noticeable indicator of
group identity. In the history of nations, especially in Europe, the sur¬
vival of a nation’s language has frequently been equated with the
continued existence of the nation itself. The belief in this equation is
often shared by both the nations which constitute the majority within a
state and those which form a minority. Since majority nations are
usually in a politically superior position, they can afford to be more
aggressive in the expression of their convictions, thus posing a threat
to the survival of the minority language(s).
Nineteenth-century Europe was full of cases of larger nations im¬
posing their language on smaller ones, irrespective of whether or not
the latter were prepared to change their feelings of national loyalties.
The systematic suppression of Welsh in Wales, for example, led not
only to a decline of the language, but also to the weakening of national
sentiments. When Prussia became the leader of the new Germany in
1871, its centralization policies meant the suppression of the rights of

minorities such as the Poles in East Pmssia and Poznan and the Danes
in Schleswig-Holstein, which had until then enjoyed cultural auto¬
nomy. In Alsace, which had been taken from the French in 1871 after
the Franco-Pmssian War, German efforts were geared towards trying to
turn Alsatians into convinced Germans. Many Germans thought that
this was a perfectly legitimate undertaking, since Alsatians, after all,
spoke a variety of German. The fact that, for historical reasons, Alsat¬
ians felt more akin to French people was not accepted - not in 1871,
nor again in 1940, when Alsace-Lorrraine was, once more, annexed by
German forces. The famous German sociologist Max Weber recalled a
visit to Colmar in Alsace which prompted him to reflect on the inter¬
play between national identity and history:
‘The reason for the Alsatians not feeling themselves as belonging to the
German nation has to be sought in their memories. Their political destiny
has taken its course outside the German sphere for too long; for then-
heroes are the heroes of French history. If the custodian of the Colmar Mu¬
seum wants to show you which among his treasures he cherishes most, he
takes you away from Griinewald’s altar to a room fdled with tricolours,
pompier and other helmets and souvenirs of a seemingly most insignificant
nature; they are of a time that to him is a heroic age.’
(Quoted in Gerth and Mills 1948: 176)
X.
Another example of linguistic nationalism, mentioned by Stephens

(1976), is Hungary’s change of language policy in the last century.
Before 1840 the Hungarian state encompassed Hungarians, Slovaks,
Germans, Romanians, Croats, Ruthenians and Slovenes. For nearly a
hundred years they had coexisted in the kingdom and, one must as¬
sume, found some common identity in its traditions. Latin was the
official language of government, but the vehicle for trade, education
and science was German. In 1844 Hungarian replaced Latin and the
way was prepared for a monolingual Magyar state. However, the pol¬
icy of magyarization at all levels infringed the linguistic rights of those
citizens who spoke languages other than Hungarian (although German
fared better than others). The concept of a multilingual and multina¬
tional state had clearly been abandoned, presumably for the sake of
creating a more powerful nation-state (and perhaps also in order to
secure the dominance of the Magyars).
The twentieth century, too, has seen many instances of language,
politics and nationalism intermingling in the affairs of countries. The
First World War was fought, among other things, for the rights of small
nations, and it was the intention of the Treaty of Versailles (1919) to
solve many of the problems that smaller nations had experienced by
implementing the principle of self-determination. Thus a number of
nations, e.g. Poland, Finland, Estonia, Latvia, Lithuania and Czechos¬
lovakia, gained independent status. The Turkish and Austro-Hungarian
empires were dissolved, and autonomous status was granted to some of
their component nationalities. Language played a prominent part in ar¬
riving at these decisions, because many of the nations in question
defined themselves in terms of their languages. However, it proved im¬
possible to satisfy the aspirations of all the nationalities involved, and
several of the newly created states were themselves ethnically and lin¬
guistically diverse.
Also today we find plenty of examples in the world where language
assumes a political role. The 1960s saw considerable civil unrest and
militant activity in Canada, where members of the French-speaking
province of Quebec demanded separation from the rest of the state. In
Belgium many governments have been brought down because of their
inability to satisfy the demands of the two main national groups, the
Flemings and the Walloons. Although further major linguistic confron¬
tations have probably been averted by the substantial constitutional
changes that have now been made, Belgium was in the news again in
1986 when the inhabitants of a little-known area called Les Fourrons
voiced their grievances about linguistic borders. In Wales, too, nationa¬
lists have frequently been prepared to resort to such acts as daubing or
taking down road signs, or even burning holidays homes owned by
English people. Their fight for a separate television channel in Welsh
(which they saw as beneficial for the language) was won from the
British Government only after the leader of the Welsh party. Plaid
Cymru, who was a well-known pacifist, had vowed to fast to death if
the demand was not granted.
Of course, it must be recognized that many of the problems men¬
tioned above can be attributed to sociocultural and socio-economic
causes, i.e. that such questions are not purely linguistic in nature or in
their origins. But language issues seem to be highly sensitive, and the
way people talk about their language does show that they attach a great
deal of importance to it. People do not normally think or talk much
about language, as it is something one takes for granted. But when we
feel threatened or provoked in our language use we may express our
feelings towards our own tongue in highly emotional terms. On such
occasions language is identified as an important element of one’s na¬

tionality, as a common heritage that provides the link, not only with
one’s own forefathers, but also with the group’s history. Fishman
(1972b) points out that language serves as a link with ‘the glorious
past’ and with the authenticity of the group. Others have referred to
their language as the living memory of their history.
Within the context of nationalism, we can discern two different
functions of language: (1) it provides for its speakers a link with then-
past and with their fellow-speakers, and therefore it helps to maintain
the group s feeling of unity; and (2) it helps to set the group apart from
others. These are actually the two sides of the same coin. Fishman
(1972b) refers to them as ‘contrastive self-identification.’ Garvin and
Mathiot (1970: 369), in their more general discussion of the roles of
standard languages, talk about ‘the unifying and separatist functions’.
The existence of a national language may provide a nation with
welcome ammunition in its fight for independence or at least some
form of autonomy. The Catalans, for example, have taken this last
route (autonomy) and been successful, on the basis that a majority of
people in the area speak Catalan. In Wales, by contrast, less than one-
fifth of the population speak Welsh, and among them only the Welsh
nationalists would advocate full autonomous status for the region. As
mentioned earlier, we cannot automatically equate language with na¬
tion. There are many nations that share the same language, e.g. the
Canadians, Americans, Australians and British (among many others),
but could not be said to have the same nationalist feelings. Similarly,
Spanish and Mexican, or French and French-Canadian, nationahsms
are of a totally different nature. It is equally inappropriate to define
nationalism purely in terms of language. In Alsace and Luxembourg it
is the regional spoken varieties of German, Alsatian and Letzebuerg-
esch, that form part of the national identity, not Standard German,
although the latter is used as the written medium.
A language may contribute, then, to shape and maintain feelings of
national identity and solidarity. But in the origin and preservation of
such feelings there are other aspects that are just as important: shared
ethnicity and history, religion and colour-consciousness, too, to name
but a few. Language may also play a central role in separatist move¬
ments, as in the emergence of new states, and in the maintenance of
power in multilingual nation-states.
10.1.3 The political dimension of language

The idea that there is a strong link between language and nationalism
and, consequently, that language should be the basis for political deci¬
sions (in the wider, programmatic sense), is a thoroughly European
one, even though it has subsequently spread outside this continent. It
emerged after the Reformation, when Latin ceased to be used as a
lingua franca and most countries saw the standardization of the lan¬
guage of the dominant national group. ^ The foundation of bodies
concerned with the preservation and purity of linguistic standards need
not be seen, in itself, as an act of linguistic nationalism, but as reflect¬
ing a preoccupation with matters of language and, perhaps, also pride
in one’s national language.
Language became an important political issue later in the eighteenth
century. In Spain the centralist policies of the new Bourbon dynasty
resulted in the systematic suppression of the indigenous minority lan¬
guages (Basque, Galician and Catalan). In France the supremacy of
French over the other regional languages was already advocated under
the ancien regime. But the Jacobins (radical revolutionaries), at the
time of the French Revolution (1789-99), were far more rigorous in
their approach to fostering linguistic uniformity. By proclaiming the
ideal of ‘une langue, une nation’ they openly linked language and poli¬
tics, claiming that national unity under a new democratic system could
be achieved only by abandoning the regional languages of France.
In Germany the rationale for the link between language and politics
was provided by philosophers such as Herder (1744—1803), Fichte
(1762-1814) and Wilhelm von Humboldt (1767-1835). Herder’s ideas
about language and nationalism developed from his philosophical work
on the origins of language, contained in his essay ‘Uber den Ursprung
der Sprache’ (‘On the origin of language’), for which he received a
valued prize from the German Academy of Arts and Sciences in 1770
and which was published in its final version in 1789. The political role
of language was implicit in Herder’s argument that language is an im¬
manent part of the development of individuals and groups that speak it
and, as such, it expresses the consciousness of groups and therefore
constitutes the most important and precious possession of its speakers.
More simply, language is the mirror of a nation. The relationship be¬
tween language and national identity was thus established. Fichte took
Herder’s general (and neutral) ideas about the nature of language fur¬
ther, pronouncing value-laden judgements on specific languages in a
nationalistic manner which would be unacceptable today. For Wilhelm
von Humboldt, the spirit of a nation was best expressed by its language
(indeed, ‘its spirit is its language’); the possession of a language with a
long history is, consequently, the most valuable asset that a nation can
have.
Nowadays language can still be a powerful factor in nationalism,
both at the level of theoretical discussion and when ideas are translated
into political action. The language policies adopted in Hungary, France
and Spain have served as historical illustrations. Events in eastern
Europe, where an increasing number of nationalities (Latvians, Ukrai¬
nians, Moldavians, Slovaks, etc.) demand greater political and
linguistic independence, provide vivid examples of the late 1980s. The
history of the world is full of instances of nationalism (turned into
imperialism) leading to human misery and cultural destruction, as
people have frequently trespassed on the linguistic and geographical
territory of others in the name of national unity. Yet nationalist feelings
have also proved to be a positive force in cases where they have kept
together groups of people which have felt threatened by other groups.
At times, they have helped to ensure the survival of a national lan¬
guage (e.g. in Catalonia, which constitutes one of the most impressive
examples in present-day Europe - see Chapter 13). Whether one sees
nationalism as good or evil, rational or irrational, there is no doubt that
in the past it has proved to be a powerful force in the world, as is still
the. case today. Chapters 11-14, which deal with linguistic minorities in
general and specific multilingual settings, further illustrate the import¬
ance of these issues.
10.2 Language planning
One of the most important decisions an emerging new nation (or newly
autonomous region) has to make is the selection of its national lan¬
guage (or its official regional one). Choosing a flag and a national
anthem usually presents relatively few problems. The making of a cons¬
titution may be a more difficult undertaking. But the choice of a lan¬
guage that is to have symbolic function and practical application for a
nation can be a complex issue with far-reaching consequences. Ideally,
the chosen language should also serve as official language, i.e. be used
as the medium of internal communication. We shall look now at the
role language plays in nation building and in the development of a
nation, and will conclude the chapter by dealing with some practical
LANGUAGE PLANNING 205
considerations raised by language policies.

Nowadays very few nations actually find themselves in the position
of having to select a national language. But there are many countries
attempting to change from one particular strain of language to another,
or trying to change the assignment of particular functions from one
language or variety to another. Some countries are reconsidering their
language policies with a view to establishing more than one national
language or more than one official language, while others find them¬
selves forced to provide administrative or educational services for one
or several of their linguistic minorities. These decisions require, in
some instances, the linguistic systems themselves to be the subject of
close scrutiny. When this happens, the language(s) may be found not to
be adequate for particular communicative purposes, or to be in need of
elaboration or standardization. Language planning is concerned both
with the symbolic function of language within a society and with the
instrumental use that its speakers make of their language.
In the following section two main areas of language planning will
be discussed: (1) language development, which centres on the language
itself; and (2) language determination and allocation of language use,
which deals with language selection and the attribution of status to
particular languages.
10.2.1 Language development

Kloss (1968) differentiates two basic aspects of language planning by
distinguishing between corpus planning and status planning. The for¬
mer refers to changes made in morphological or syntactic structure,
vocabulary or spelling (the ‘corpus’); it may even include the adoption
of a new script. The purpose of ‘status planning’ is to develop a lan¬
guage or a variety of a language, with a view to making it a useful
instrument for communication, i.e. one that can fulfil all functions
within society. Ferguson (1968) speaks of language development rather
than corpus planning, and he describes its three basic stages as graphi-
zation, standardization and modernization.
Graphization involves taking decisions such as what kind of al¬
phabet is to be adopted for a language that has not yet been reduced to
writing, and what the conventions should be for spelling (including
accentuation if appropriate), punctuation and capitalization. In estab¬
lished nations the development of a particular writing system usually
spans a long period of time, and it can involve changes which, at the
time when they are made, may be quite small.^

Standardization refers to the selection of one variety which is con¬
sidered to be ‘the best’ and is to serve as the norm for the speech
community. Ferguson refers to standardization as a process of becom¬
ing more and more widely accepted throughout a given society, thus
emphasizing the fact that, usually, a standard variety emerges grad¬
ually. The absence of a standard, however, often appears to be an
obstacle for the survival of a language, at any rate in twentieth century
Europe. This is certainly the case in the Romansch-speaking areas of
Switzerland. Some attempts are being made at standardization with re¬
gard to Romansch, but they may have been started too late (Schlapfer
1982).
Modernization means the development of a variety so that it will be
suitable as a vehicle for communication able to fulfil all functions. A
process of modernization is one during which the linguistic resources
of the language are increased, so as to make it a more useful tool for
its users. This may imply extending the lexicon, either by coinage or
by borrowing, and introducing new expressions. Vocabulary expansion
normally suffices for the introduction of new ideas, products, technical
terms, etc. Sometimes it is also undertaken in the hope of ridding the
language of foreign influences.^
Modernization can also be applied to a spelling system, either partly
or totally. For example, in the 1920s when Ataturk set about estab¬
lishing a modem Turkish state, he changed the writing system from the
Arabic script to the Roman. Similarly, the USSR had to take a decision
on writing conventions within its vast multilingual and multigraphic
state. Modernization there involved, among other things, the extension
of the Cyrillic script to nearly all the languages of the Soviet Union. In
China language development has taken the form of trying to make
Chinese easier to write - the aim being to simplify about half of the
estimated 7000-8000 characters. Although the simplified characters
should, one would think, really benefit everyone, an important obstacle
is that they are employed only for materials printed in China itself - in
Hong Kong, Taiwan and other Chinese-speaking communities the
traditional characters continue to be used. Simplification in one area,
then, may lead to splitting up the language into divergent forms, per¬
haps reflecting political divisions. In general, language planners may
make recommendations, but ultimately language use is subject to un¬
controllable and even, at times, irrational forces.
10.2.2 Language determination and allocation of language use

Language development (or ‘corpus planning’) and language determina¬
tion (or ‘status planning’) can be treated as theoretically separate
notions, but in practice they are interlinked. The involvement of the
Welsh Language Society or the Catalan autonomous government in
language development, for instance by expanding the lexicon, is also a
reflection of the fact that certain decisions have been taken at a politi¬
cal level - namely, to use Welsh, or Catalan, in schools, as a medium
of instruction, or in local administration. But what prompted these de¬
cisions? How were they reached, and what is their ultimate aim? The
section of the discipline of language planning that attempts to address
such questions has received a good deal of attention in the last twenty
years or so, as it concerns decision-making processes regarding the
status and function of particular languages or varieties, as well as the
allocation of state resources. Various labels have been proposed for
this area of study. Neustupny (1970; 4) speaks of a ‘policy approach’,
whereas Jemudd (1973: 16-17) discusses ‘language determination’.
Kloss’s (1968) concept of ‘status planning’ has already been men¬
tioned. Rubin (1983: 341) suggests that the term ‘allocation of
language use’ would be more useful, as does Cobarrubias (1983: 42),
who refers to ‘allocation of language functions’ for a language in a
given speech community. The areas dealt with do not correspond
exactly with each other in coverage or in approach, but they may be
loosely grouped together for convenience.
In his 1968 study Kloss uses the term ‘status planning’ to refer to
those planning decisions intended to enhance or diminish the status of
a language. He establishes four categories that relate to language
status.
(1) The origin of the language used officially. It may be an indigenous

language, i.e. one spoken by a large part of the population, in which
case the state is called an endoglossic state. Most European countries
are examples of endoglossic states. If, on the other hand, the na¬
tional/official language is an imported one, we speak of an exoglossic
state. Many of the states that used to be colonies, which may be lin¬
guistically heterogeneous, have chosen the former colonial language as
their national one, usually for purely practical reasons; examples are
Ghana, Nigeria, the Cameroons and some of the smaller South East
Asian countries, e.g. Malaysia, East and West Samoa, and the Philip¬
pines. In order to accommodate those states which are neither
endoglossic nor exoglossic but show features of both, the term mixed
states is used. They include, for instance, states that are in the process
of shedding the colonial language in favour of an indigenous one, like
Papua New Guinea (which has three official languages; Hiri Motu,
Neo Melanesian or Tok Pisin, both pidgins, and English), or Tunisia
(which also has three languages: Arabic, Tunisian and French).
(2) The developmental status of a language is the second criterion for

Kloss. According to the degree of elaboration, i.e. as a result of lan¬
guage development, a language may achieve one of six types of status:
(a) a fully modernized standard language that fulfils all needs of

modem society, e.g. French;
(b) a standard language spoken by a relatively small group only
which, because it is not used by many speakers, has limited
scope, e.g. Welsh;
(c) an archaic language, which once flourished but is not now
equipped for coping with modem science and technology, e.g.
Latin;
(d) a young standard language, recently codified and standardized
for some specific purpose (e.g. for education), as for instance
Luganda in Uganda;
(e) an unstandardized language for which alphabetization has only
recently been carried out, as has happened with Somali in Somalia;
(f) a preliterate language, i.e. one which has not yet undergone
graphization, e.g. Gallah in Ethiopia.
(3) The third category of language status considers languages with re¬
gard to their legal position. As a result of language-planning decisions,
a language may be recognized as:
(a) the sole national official language, such as French in France or

German in Germany or Austria;
(b) a joint official language, i.e. co-equal in terms of use for govern¬
ment functions, as happens with French and Flemish in Belgium,
English and Afrikaans in South Africa, or French, German and
Italian in Switzerland;
(c) a regional official language, i.e. one which enjoys official status
on a regional basis, like Catalan in Catalonia, German in eastern
Belgium, or Ukrainian, Georgian and Armenian (among others in
the Soviet Union);
(d) a promoted language, which is one that is used by various auth¬

orities for specific purposes, as is the case with Spanish in some
parts of the United States or with West African Pidgin English in
Cameroon;
(e) a tolerated language, i.e. one that is neither officially supported
by public agencies nor proscribed; there are many languages in
western Europe which are more or less left to their own devices,
e.g. Basque and Catalan in France, the Asian languages of immi¬
grants in the UK, or Turkish, Greek, Italian, Spanish and other
migrants’ languages in western European countries;
(f) a proscribed language, i.e. one against which there exist govern¬
ment restrictions and sanctions; this was the position of Catalan,
Basque and Galician in Franco’s Spain, of Welsh and Irish
Gaelic until the 1920s in the United Kingdom, or of Scots Gaelic
after the uprising of 1745.
(4) The ratio of users of a language to the total population is Kloss’s

last factor among those that can affect language status. He does not try
to establish any kind of quantification-based criteria. The question of
ratio alone does not actually say much about the status of a language;
for this reason, he suggests, this factor should always be considered in
conjunction with others.
These categories, however, do not take into account the area of lan¬
guage function, i.e. the uses to which it is put (for instance, in
education, or in religious worship), which is today considered to be
highly relevant in determining language status. Experience has shown
that the prestige of a language is enhanced significantly, for example,
once it becomes accepted as the medium of education. The adoption of
Swahili in Tanzania as a language of instmction up to secondary level,
or that of Quechua in Peru or Welsh in Wales, contributed to making
these languages more acceptable, apart from gaining more speakers for
them. The decision by the Catholic Church in favour of the use of
vernacular languages instead of Latin provided a substantial boost for
minority languages (e.g. in American Indian communities in Central
and South America) and also for Creole varieties, which began then to
be used in religious services in place of the standard forms.
Kloss, then, sees status planning primarily as resulting from a series
of planning decisions and in terms of official attitudes towards a
country’s language or languages. But language status can be viewed
from a different perspective. Mackey makes the point that the status of
a language depends on the individual and ethnic circumstances of the

people and the languages that they use:
The status of a language depends on the number of people using it, their
relative wealth, the importance of what they produce, their social cohesive¬
ness and the acceptance by others of their right to be different. In other
words, the faces of language status are demografic, economic, cultural, so¬
cial, political and juridical
(Mackey 1983: 174).
10.2.3 Language-planning policies - some examples

North America Mackey, when he wrote the above, was comparing Ca¬
nadian language policies with those of the United States. He makes the
point that a number of notable differences exist between the French-
speaking population in Canada and the Spanish-speaking groups in the
neighbouring country, which would make the adoption of the Canadian
language planning policies unworkable in the United States. The main
differences can be seen in the fields of education and the law. In the
United States various types of bilingual education programmes invol¬
ving Spanish and English have been adopted over the past twenty
years or more. For a number of reasons, these programmes have not
been very successful, in particular at the higher levels of education,
where there is no provision for Spanish at all. In Canada bilingual
education has also encountered problems, mainly outside the French-
speaking province of Quebec. In Quebec French-language education
has been available for a long time and at all levels, and this has pro¬
moted the teaching of French to children of French-speaking minorities
outside Quebec. There are now also bilingual education programmes
providing for the teaching of French to the English-speaking majority.
These are measures that have no counterpart anywhere else in Ameri¬
ca.
The legal status of a language will, of course, to some extent deter¬
mine what kind of education programmes a state may embark upon.
The status of Spanish in the United Sates is based on two pieces of
legislation, the Voting Rights Act of 1965 and the Bilingual Education
Act of 1968, but neither of them officially accords Spanish ‘official
status’ - they just pave the way for Spanish to be used, or officially
supported, for specific purposes such as in education, for voting, in the
courtroom and in the media.
In contrast, after two centuries of struggle and as a result of active
participation in the political decision-making process of Canada, the

French-speaking population has achieved official status for its language
(Languages Act of 1969). There followed a series of major legislative
reforms that first changed Canada from an officially monolingual to an
officially bilingual state, and subsequently Quebec from a bilingual
area to a monolingual one with French as the official language. In lan¬
guage-planning terms, these reforms represented a reallocation of
language functions. But the repercussions of these language policies go
beyond linguistic matters, as they affect social planning as well. The
Act to Promote the French language in Quebec (Bill 63 of 1969) af¬
firms the rights of parents to choose either French or English as the
language of education for their children; and it provides for children
from non-French backgrounds to acquire a working knowledge of
French. The Official Language Act (Bill 22 of 1974) establishes French
as the official language of Quebec; it declares that the French text of
Quebec’s statutes must prevail over the English version in cases of
disagreement, thus defining the official character of French. It allows
for bilingualism but specifically states that professional bodies and
public agencies must provide their services in French, just as all offi¬
cial texts must be printed in this language. The Charter of the French
Language (Bill 101 of 1977) enacts the shift from bilingualism to
monolingualism, by declaring that only the French texts of all state
publications are official, and it requires every public utility and busi¬
ness firm to use French in their publications (brochures, catalogues,
order forms, job application forms, etc.) and to obtain a ‘francization
certificate’ which attests that the firm is applying a ‘francization pro¬
gramme’. All public education at primary and secondary level is to be
in French. A number of supervisory bodies are set up to monitor the
implementation of the Charter. It is easy to see in this case how a
language-planning decision can directly affect people’s lives. There is
no doubt that the Charter represents a major piece of planning de¬
signed to enhance the status of French in Canada.
Spain The Spanish Constitution of 1978 is also interesting in this
context, as not only does it make provision for the official language,
but it refers also to the status of the other languages of Spain which,
for much of Spain’s history, were not officially recognized. Thus Ar¬
ticle 3 of the Spanish Constitution states that ‘Castilian Spanish is the
official language of the Spanish state’ and that ‘all Spaniards have the
duty to know and the right to use it’. As regards Catalan, Basque and
Galician, it acknowledges in the next paragraph that ‘the other Ian-
guages of Spain will have co-official status in their respective autono¬

mous communities as laid down in their Statutes’ (of Autonomy),
affirming that ‘the richness of the different linguistic varieties of Spain
is a cultural heritage that must be the object of special respect and
protection’. When the Generalitat drew up its own Statute of Auton¬
omy for Catalonia it gave the language question considerable weight,
stating in its Article 3 that:
Catalonia’s own language is Catalan;
The Catalan language is the official one in Catalonia, together with Castil¬
ian Spanish, which is official throughout the Spanish state;
The Generalitat will guarantee the normal and official use of both lan¬
guages, will take the necessary measures to ensure the knowledge of them
and will create conditions which will allow them to attain full equality with
respect to the rights and duties of the citizens of Catalonia.
The first sentence reflects the link between nation and language,
which has always been perceived by Catalans. It also seems to confirm
an underlying territorial principle which implies that Catalan is to be
used ‘in Catalonia’, not just ‘by the Catalans’; and it paves the way for
subsequent legislation (the Law of Linguistic Normalization of 1983)
to aim at making Catalan the language normally used in all walks of
life, particularly in the fields which can be more directly influenced by
the regional government, namely education and administration. The
second statement declares the official status of Catalan. The third part
of the Article is a guarantee of bilingualism, while at the same time it
provides for the promotion of Catalan (since it can be taken for granted
that everyone knows Castilian Spanish already).
The Catalans were well prepared for language planning when they
finally found themselves in a position to do something about Catalan.
They had taken on board some of the experiences of other countries (in
particular Wales and Canada), and they understood the limitations of
officially sponsored bilingualism: only a considerable reallocation of
language functions from Castilian Spanish to Catalan would, in their
view, ensure a degree of success in their efforts to halt the decline of
Catalan (see Chapter 13).
United Kingdom Language policies are seldom formulated and earned

out in one fell swoop. Quite often they are not officially stated at all,
and language is allowed to develop without any kind of planning, car-
ried along by social and political forces. The development of English

as the de facto national (and official) language of England took over
200 years; for decades after the Norman Conquest most political deci¬
sions were taken by men who spoke Norman French and used it in
administration. Individuals, rather than governments or institutions,
were responsible for enhancing the status of English, thus contributing
to its eventual adoption as the language of the government and of Eng¬
lish society (for an account see Heath and Mandabach 1983). Any
form of legal regulation or official involvement with language has been
singularly absent in British history. In the United Kingdom there is no
law proclaiming that English is the national language, as there is no
written constitution, and no official body, such as an ‘Academy’, en¬
trusted with looking after the language. There have been, however,
instances of negative language policies, such as the suppression of
Scottish Gaelic, Irish and Welsh, or the imposition of English, to the
exclusion of the official use of the vernacular, in the colonies. Perhaps
one of the most extreme statements of linguistic chauvinism is Thomas
Macaulay’s ‘Minute on Education’ of 1835. Macaulay was a member
of the Supreme Council of India and he became involved in the debate
over the medium of instruction in Indian schools. It was his idea that a
class of Indians should be educated to acquire English moral and intel¬
lectual values, so that they could act as interpreters between English
rulers and those ruled by them. On the question of which language
should be used he wrote:
The claims of our language it is hardly necessary to recapitulate. It stands

preeminent even among the languages of the West. . . It may safely be
said that the literature now extant in that language is of greater value than
all the literature which three hundred years ago was extant in all the lan¬
guages of the world together. . . The question now before us is simply
whether, when it is in our power to teach this language, we shall teach lan¬
guages in which, by universal confession, there are no books on any
subject which deserve to be compared with our own.
(Quoted in Edwards 1985: 31)
In the 1920s important steps were taken in Wales and Ireland:

Welsh became an officially accepted (but not promoted) language and
Irish was chosen as the national language of the new Republic of Ire¬
land. The latter was a particularly significant choice, as only a fraction
of the population spoke Irish - yet a majority of the people, for reasons
of nationalism, felt that Irish Gaelic was a better symbol of their ident-
ity than English; also, demands to preserve, and if possible revive, the
Irish language had come from respected quarters. But the Irish Con¬
stitution spells out that English is the second official language, and
today virtually everyone in Ireland uses English for most functions.
More recently, language policies have been initiated at a local level
with a view to making some educational provision for Britain’s new
linguistic minorities. This has primarily involved teaching English as a
second language to children who started school with little or no knowl¬
edge of it. Some local education authorities now also provide
mother-tongue classes to non-native English speakers, although often
in a rather informal, unsystematic manner. The initial impetus, inciden¬
tally, was due to EC legislation (see Appendix A).
10.2.4 Practical considerations

Language policy decisions can be put into practice in a variety of
ways. A government may formulate laws which specify the use of a
particular language, for instance in government administration; or it
may encourage the world of business and commerce to adopt certain
linguistic practices by offering subsidies or threatening sanctions; and
publishing may be sponsored in one particular language and/or re¬
stricted in another. The education system is by far the most important
tool for implementing a government’s language planning policy.
Government policies can also be carried out indirectly, in the sense
that they may not be deliberately set out by the government but never¬
theless reflect official attitudes. For instance, in the last 200 years or so
the attitudes of the state towards minority languages often came to the
fore in the armed forces. In countries with military conscription the
rule was usually to insist on the use of the national language within all
sections of the services. Consequently, speakers of regional languages
or varieties, for example in France and Spain, had to learn to use the
official language or at any rate to improve their knowledge of it; in
Italy and Germany dialect speakers had to conform to the standard
variety. The armed forces (through the system of compulsory military
service) thus played a part in promoting the official language and in
suppressing regional variations.
At a non-govemment level, political parties and other organizations
interested in promoting a language, such as the Welsh Language So¬
ciety, can put pressure on governments to take actions that favour their
language, and they can organize language classes or engage in cultural
activities in the languages concerned, as does the Cercle Rene Schi-

ckele in Alsace.
There are many different reasons why deliberate language planning
is undertaken. Political, educational and practical considerations may
lead planners to formulate policies that have the effect of changing the
status of a particular language or variety. Cobarrubias (1983) argues
that certain planning tasks are not ‘philosophically neutral’ and that
moral issues are often involved in planning. One such issue concerns
the question of language rights: do immigrants give up their right to
use their language and keep their cultural identity when they choose to
migrate, or is it the right of people to retain and maintain their lan¬
guage wherever they are? The United Nations and UNESCO have
declared that ethnic groups do have the right to maintain their lan-
guage(s), but few countries have incorporated this principle into their
legislation. Directive 486/EEC (see Appendix A), issued by the Euro¬
pean Community in 1977, confirmed the right of migrants’ children to
receive mother-tongue teaching. This directive goes some way towards
recognizing certain language rights. It has not, however, been a very
effective measure, as its application was left to the discretion of each
individual country.
Cobarrubias (1983: 63) suggests four typical ideologies that may
motivate language planning:
(1) linguistic assimilation;

(2) linguistic pluralism;
(3) vemacularization;
(4) internationalism.
Linguistic assimilation is the name given to the notion that all inhabi¬
tants of a country should speak the dominant language, regardless of
their origin. There are numerous examples of states which have ad¬
hered, consciously or unconsciously, to this belief. A corollary of this
position is that linguistic superiority is attached to the dominant lan¬
guage, and that minority languages are not, per se, granted equal
rights. The United States has practised linguistic assimilation widely
(and still does) in areas which it has annexed, colonized or acquired,
such as Hawaii, New Mexico, the Philippines and Alaska.The case of
the hellenization of Macedonia in 1912-13 can be seen as a European
example of a country that has indulged in the same kind of policy.
Linguistic pluralism, roughly speaking, recognizes the coexistence of

different language groups, which are granted the right to maintain their
languages. There are various forms of pluralism, depending on the
status of, and the official support given to, the individual languages.
South Africa, Singapore, Belgium, Switzerland and Finland are exam¬
ples of countries where two or more languages are officially
recognized.
Vernacularization concerns the restoration or development of an in¬

digenous language or variety, and its adoption as an official language.
The promotion of Quechua in Peru and of Tagalog in the Philippines
are examples of this.
Internationalism involves a country choosing a language of wider com¬

munication (rather than an indigenous one) as an official language, or
as one to be used for education purposes. The choice of English in
Kenya, India, Singapore and Papua New Guinea illustrates this type of
ideology.
The underlying ideology and a country’s interpretation of language

rights will therefore shape its language policies. Planning decisions, in
turn, will often determine the fate of minority languages, thus reflect¬
ing official attitudes towards minorities. Cobarrubias (1983; 71 ff)
discusses a number of these attitudes (but it is worth remembering that
he was not trying to present a complete taxonomy):
(1) attempting to kill a language;

(2) letting a language die;
(3) unsupported coexistence;
(4) partial support of specific language functions;
(5) adoption as an official language.
Of course it does happen that official attitudes change over the

years. For instance, in Spain government policy towards Basque has
shifted from total suppression (‘attempting to kill’) in the years after
the Spanish Civil War, through ‘unsupported coexistence’ in the last
years of Franco s dictatorship, to official recognition and regional sup¬
port in present-day deitiocratic Spain.
Language planning is also an ongoing process in older nations in
Europe, where the question of language status seems to have been de¬
cided long ago. But changes need to be made from time to time as
linguistic minorities, both the indigenous and the non-indigenous ones,

become more articulate in formulating their demands. States that used
to see themselves as monolingual now find themselves compelled to
address the question of language allocation anew. Finland, Sweden and
Norway, for example, have already made some changes with respect to
the status of Same (or Sami), and the Netherlands have allowed Frisian
to be used as the medium for instruction in schools in Friesland; Ger¬
many and the Scandinavian countries are making special provisions for
the education of the children of migrants. In Canada, as we have seen,
the emphasis is not only on supporting bilingual education, but on real¬
locating a language (French) in the domain of business and commerce,
where it played little part before.
The United States, too, provide an interesting example of language
planning measures being taken in response to a changing linguistic
situation - and also of public reaction to them. By far the biggest wave
of immigrants in recent times (particularly from the 1960s onwards)
consisted of Spanish-speaking people, followed by speakers of Asian
languages. By 1984 an estimated 17.5 million (Wardhaugh 1987) His¬
panic people had settled in the country and, contrary to what happened
with previous immigrants, had shown little sign of being in the process
of a.s.similating linguistically to American English. The 1968 Bilingual
Education Act and subsequent legislation had been passed to promote
bilingual and mother-tongue provision in various fields of public life.
At the beginning of the 1980s, however, public opinion had turned
against bilingualism, and demands for curbing bilingual education pro¬
grammes and other facilities were being frequently voiced. The peak of
reaction was reached in 1981 (and again in 1983) when proposals were
put before Congress for an amendment to the Constitution so as to
declare English the official language of the United States. The moves
did not succeed. If such proposals had been accepted, the federal go¬
vernments would have been prevented from maintaining, as a general
entitlement, such pluralist policies as were reflected in the bilingual
education schemes and the decision to publish official documents (e.g.
voting papers) in several languages. The debate about language rights
in the United States has not subsided, and some individual federal
states have actually proceeded to declare English the official local lan¬
guage (for a detailed discussion, see Wardhaugh 1987; Marshall 1986).
Notes
1. Before the end of the eighteenth century most of the major Euro¬
pean countries had produced authoritative grammars, including
spelling rules, and famous writers, philosophers and scientists, by
using the national languages, were enhancing their prestige. In
many countries national academies were established and charged
with the task of maintaining linguistic standards. The Accademia
della Crusca in Florence was founded in 1582, the Academie
Frangaise in 1635, the Deutsche Akademie der Wissenschaften
(Berlin) in 1700, the Real Academia Espahola de la Lengua in
1713 and the Russian Academy in 1724, to name some of the
major ones.
2. One example would be the dropping of the final ‘me’ in words
like programme’ in some forms of contemporary English, an¬
other the decision to remove the written accent from
monosyllabic words such as ‘fue’, ‘fui’, ‘vio’, ‘dio’ in Spanish,
taken by the Royal Academy in the 1950s. The German Federal
Republic decided, after the Second World War, to use the
Roman alphabet and to give up printing and writing in the Go¬
thic script. In other cases, as for instance in Indonesia, the
decisions to adopt the widely-used pidgin (Bahasa Indonesia) as
official language meant that graphization covering a whole lan¬
guage had to be carried out within a short period of time.
3. There have been many such attempts - and almost as many

failures, e.g. the Sprachgesellschaften’ (learned societies con¬
cerned with the protection and furtherance of the German
language) in eighteenth- and nineteenth-century Germany, trying
to fend off French influences, Ataturk’s efforts to rid Turkish of
Arabic and Persian loanwords, Israel’s desire to expand modem
Hebrew by means of new coinages rather than borrowing words
from European languages, and the exertions in present-day
France to resist the increasing use of English words in ordinary
people’s use of French; one typical example was the short-lived
attempt, in the early days of Franco’s rule in Spain, when foot¬
ball was becoming a popular spectator sport, to coin the word
‘balompie’, which was considered to be authentically Spanish,
for people to use instead of the loanword ‘futbol’, which is now
firmly established in the language.
Chapter 11
Linguistic minorities
The basic concepts introduced in Chapters 8, 9 and 10 are exemplified

in the rest of the book. The discussion in this chapter deals, in the
main, with general aspects of the subject of linguistic minorities and
issues that are relevant to language use and language maintenance
within these groups, as well as the factors which bear upon their conti¬
nued existence as separate communities. The focus of attention is the
linguistic minorities of western Europe (the analysis of the rich linguis¬
tic diversity and complex patterns of language use in central and
eastern Europe falls outside the scope of these chapters). The case
studies that follow later (Chapters 12 to 14) present a more detailed
consideration of three very different sets of situations in which linguis¬
tic minorities in western Europe today find themselves. It is therefore
hoped that the present chapter illustrates the theoretical points made
previously and that the case studies will draw together the various
strands of political, socio-economic, cultural and educational issues
that have been analysed in preceding chapters.
A look at a map of western Europe (including the north and the
south) reveals an astonishingly diverse picture: of the eighteen coun¬
tries included in it (Greenland, Iceland, Finland, Norway, Sweden,
Denmark, Germany, Austria, Switzerland, Belgium, Netherlands, Lux¬
embourg, France, Spain, Portugal, Italy, the UK and the Republic of
Ireland, leaving aside the very small communities of Andorra, Monaco,
San Marino, Lichtenstein and Vatican City), only two, Portugal and
Iceland, have no indigenous linguistic minorities. In some of the bigger
countries such as France, Britain, Spain and Italy, there are several
linguistic minorities, yet these countries are often thought of as mono¬
lingual. One is used to thinking that in Britain people speak English, in
220 LINGUISTIC MINORITIES
Faroe
Islanders
Galicians
11.1 Indigenous linguistic minorities in western Europe

INDIGENOUS MINORITIES IN WESTERN EUROPE 221
France French, in Spain Spanish, and so on. This happens, not so much
because the minorities are small in terms of numbers of speakers (in¬
deed, some of them have millions of them), but rather because some of
these minority languages are not perceived by the population as a
whole as having equivalent status to the national language, and there¬
fore there may be insufficient social support for them. On the other
hand, smaller countries such as Switzerland, Belgium and Luxem¬
bourg, which officially recognize the existence of other languages
within their borders, are considered multilingual, although the numbers
involved are relatively small.
11.1 Indigenous minorities in western Europe
An indigenous linguistic minority is a community of people who share

a number of common characteristics, among them the fact that they
speak their own language, and who perceive themselves as different
from the groups of speakers of the majority language. They are settled
in a given area, the minority area, where they have lived for a consid¬
erable length of time, usually for centuries. In that area the language in
question, which is not that of the majority of that state’s citizens, is
spoken by all, or at any rate a sizeable proportion, of the members of
the linguistic community. The following list (based on Stephens 1976)
mentions some of the more important linguistic minorities of western
Europe.
TABLE 11.1 Linguistic minorities of western Europe
State Minority Name of

minority language
Finland The Swedish Finlanders Swedish

The Same Same (or Sami)
Sweden The Finnish Swedes Finnish

The Same Same
Norway The Same Same
Denmark The Germans of North Schleswig German

The Faroese Islanders Faroese^

minority language
West Germany The Danes of North Schleswig Danish

The North Frisians Frisian
Netherlands The West Frisians Frisian
Belgium^ The Flemings Flemish

The Walloons French
The Germans of Old Belgium German
Luxembourg The Letzeburgers Letzebuergesch'^
France The Bretons Breton

The Flemings of Westhoek Flemish
The Occitans Occitan
The Catalans of Roussillon Catalan
The Northern Basques Basque
The Alsatians Alsatian‘S
The Corsicans Corsican
Spain The Basques Basque

The Catalans Catalan
The Galicians Galician (or Gallego)
Italy The Piedmontese Piedmontese®

The Occitans of Piedmont Occitan
The Aostans Aâldotains Provencal
The South Tyroleans German
The Romagnols Romagnol®
The Friulans Friulian
The Ladins Ladin
The Slovenes of Trieste Slovene
The Sards Sardinian
The Catalans of Alghero Catalan
The Greeks, Albanians and Greek,
Croats of the Mezzogiomo Albanian and
Croatian
Austria The Slovenes of Corinthia Slovene

The Magyars and the Croats of Hungarian
Burgenland and Croatian
INDIGENOUS MINORITIES IN WESTERN EUROPE 223

minority language
Switzerland The Ticinese Italian

The Jurassians French
The Rhaetians Romansch
UK The Gaels of Scotland Scots Gaelic

The Lowland Scots Scots
The Gaels of the Isle of Man Manx®
The Gaels of Northern Ireland Irish Gaelic
The Welsh Welsh
The Channel Islanders French
f
Irish Republic The Gaels Irish Gaelic
(Eire)
Notes
a. The case of Belgium is somewhat special, since its entire population
can be said to consist of minorities. The German-speaking minority is
small (1 per cent), but the French- and Flemish-speaking minorities
number 32 per cent and 56 per cent respectively.
b. Faroese is not recognised by the Danes as a language in its own

right, but as a dialect of Danish. The Faroese people use standard Dan¬
ish for written purposes.
c. The vast majority of Luxemburgers use Letzebuergesh, which is an¬

other Allemanic dialect, alongside French. Letzebuergesh shares many
of the features of other minority languages, e.g. until not so long ago it
was common only in its spoken form, as High German was used for
written communication. Today High German continues to be used as a
written medium, but Letzebuergesch has been standardized and is also
used for certain written purposes.
d. Alsatian is an Allemanic dialect. It is more similar to the German

dialect forms spoken in Luxembourg and Switzerland than to High
German. Traditionally, High German was used for written purposes,
but more recently Alsatian is also used in writing. (See Chapter 12 for
more details.)
e. Manx is more or less extinct now as a spoken language. It is in¬

cluded here because it is still considered to be a part of the islanders’
cultural identity. The same, although to a lesser degree, could be said
of Cornish. This language is said to have become extinct by the end of
the eighteenth century.
f. Irish Gaelic is the national language of the Republic of Ireland, but

only a small proportion of Irish people use it as a language for daily
communication.
g. Piedmontese and Romagnol belong to the Gallo-Italic group of dia¬

lects, the largest of the northern dialects in Italy.
h. Friulian and Ladin both belong to the Rhaeto-Romance language

family.
The names of the minorities refer to the language with which they
are associated, not to their citizenship. Thus, for example, the Danes of
North Schleswig hold German or dual nationality, while they see them¬
selves as belonging to the Danish-speaking community.
Those minorities whose language (e.g. French or German) is also
that of another independent state, usually speak a regional variety
(which may be quite different from the standard variety), while using
the standard form for written purposes.
In Europe historical boundaries between languages spoken natively
in each country do not, in most cases, coincide with present-day fron¬
tiers. Minorities have been the result of political decisions that have
seldom been taken on the basis of cultural, ethnic or linguistic con¬
siderations. Chapter 9 included a discussion of historical and
contemporary factors that have led to the emergence of multilingual¬
ism, i.e. new language groups were formed from the subdivision of
existing ones.
11.2 Two main types of indigenous linguistic minorities
The above-mentioned European minorities can be divided into two

broad categories:
(1) The first type of linguistic minorities are those whose language is
not the official language of any state. These minorities are ethnic
groups, mostly nations in their own right, whose ethnic homeland is in
TWO MAIN TYPES OF INDIGENOUS LINGUISTIC MINORITIES 225
the territories where they live. The French ethnologist Guy Heraud
(1963) calls this type Tes ethnies sans etat’. Examples of such self-
contained groups would be the Occitans and the Bretons in France, the
Basques and Catalans in Spain, the Welsh and Gaels in Britain, the
Frisians in the Netherlands, and the Same in north Scandinavia.
Whether or not they are recognized as nations and their languages
as national ones depends on the extent to which their speakers have
demonstrated a nationality consciousness and on the status that the ma¬
jority of the inhabitants or the host nation as a whole has afforded to
them.
(2) The second type comprises those linguistic minorities whose lan¬
guages are official ones elsewhere, i.e. communities who are in the
minority in the state of which they are citizens, but who look to their
ethnic homeland outside the state where they live. Such minorities are
usually younger than those of the first type in that they owe their exist¬
ence to comparatively recent alterations in state frontiers: for example,
the Alsatians in France, the Danes in North Schleswig and the Ger¬
mans in Denmark, the Tyroleans in Italy and the Swedes in Finland.
These minorities are free of the onus of having to prove that they are
nations in their own right. Instead, they may have the advantage that
their languages enjoy a cultural prestige out of proportion to their nu¬
merical strength. Their speakers are also able to benefit from recorded
and written material produced in the ethnic homeland, such as televi¬
sion and radio programmes as well as all forms of printed (including
educational) material. This is a very important factor for the mainten¬
ance of their language, both at a private level and in the field of
education. It is also a facility that the minorities of type (1) lack; it is
therefore much more difficult and costly for the latter to keep up their
languages.
In the case of the second type of minorities, their languages are
officially recognized. This recognition may take different forms as far
as the use of the language in public life and administration is con¬
cerned, and some form of educational provision may be made for it.
For example, German-speaking children in Tyrol are catered for in
Italy’s education system, and the same happens in Sweden in relation
to its Finnish minority (in both countries this is only a recent develop¬
ment). The interests of the Danish and German minorities in North and
South Schleswig are attended to, both educationally and in the field of
administration, with the support of the two host countries. But France
does not take Dutch into account in education or public life for its
Flemish minority.
The broad definition of ‘linguistic minority’ given at the beginning
of the chapter has been the basis of the two types just described. It
should be pointed out, though, that the term is useful only as long as
one uses it in a general way. In fact, minorities differ greatly, and in
many respects, from one case to another. They differ, above all, in the
way they see themselves and in the ways they are seen by others, and
in the kind of pattern of language use found among their members. The
Corsicans, for instance, consider themselves to be a separate national
group, but the majority of mainland French people do not recognize
their claim. The Germans in Denmark seem to share the national ident¬
ity of those of the Federal Republic of Germany, as well as that of the
Danes among whom they live, but, in contrast, the Alsatians (who
speak a Germanic dialect) do not share any feelings of belonging to the
same ethnic group with the Germans across the border, even though
they consider themselves to be ethnically different from the French.
11.3 Ethnic, regional, national and minority languages
The terms ‘ethnic language’, ‘regional language’, ‘national language’

and ‘minority language’ are all equally vague, as their connotations
depend on the political persuasion and national feelings of their users.
In addition, ‘minority language’ is unsatisfactory because it seems to
suggest that only a small number of people speak it. Yet as far as
linguistic minorities are concerned, size is very relative indeed. In
Schleswig-Holstein (north Germany) the number of Frisians who speak
Frisian is actually smaller than the section of this group of people who
do not (the same applies to the Welsh in Wales). But in South Tyrol
the group of German speakers is larger than the number of speakers of
Italian, i.e. the linguistic minority is the majority in that particular area;
the same is true for the Catalans in Catalonia, but not for Catalan spea¬
kers in the Pals Valencia next door, where they are in a minority, the
majority being native speakers of Spanish.
As we shall see later, the Catalans are unusual in many ways: there
are about six million of them who speak Catalan, which makes them
the biggest linguistic minority in western Europe and a community
larger than those of many European states, living in an area which is
more extensive than Belgium, the Netherlands or Switzerland. Catalo-
A HISTORICAL PERSPECTIVE 227
nia was mentioned earlier (p. 225) as a linguistic minority of type (1),
yet there is an independent state where Catalan is spoken as the official
language, the Principality of Andorra - which, at least theoretically,
makes it eligible for type (2) membership. Andorra may be an extreme¬
ly small country, but the fact that Catalan is spoken there means that
the Spanish government in Madrid supports interpretation into and out
of Catalan in some official transactions between the two states, a fa¬
cility which it does not grant to other languages spoken within Spain.
The case of Belgium is another one that shows up the inadequacies
of the term ‘linguistic minority’. Leaving aside Brussels, which is offi¬
cially bilingual and accounts for 11 per cent of the population, the
country is divided into two almost equal territories, Flanders and Wal-
lonia, but some 56 per cent of Belgians are Flemish speakers, 32 per
cent are French speakers and only 1 per cent speak German. Of course
there is no problem in referring to the Germans of the eastern region
(Altbelgien) as a linguistic minority, but it is more often the Flemish
rather than the French-speaking Walloons who are held to be the lin¬
guistic minority. Belgium also has over 10 per cent foreign residents,
many of whom are migrant workers and their families. On the other
hand, the Gaels in Ireland, who make up about 3 per cent of speakers,
are very much in a numerical minority; yet their language enjoys a
social prestige among the population which is out of all proportion to
the actual number of native speakers, as it is the national language of
the whole of the state of Eire. Clearly, then, size is not particularly
important in the discussion. The expression and perception of national
or ethnic identity, the status and prestige of the minority language and
its speakers within the state, and linguistic factors such as whether or
not the language has undergone standardization or is backed by a lite¬
rary tradition, as well as social aspects like the purposes for which the
language is used, appear to be more relevant.
11.4 A historical perspective
The concept of nation and its relation to language has a complicated

history in Europe, as the brief discussion at the beginning of this chap¬
ter (section 10.1) has shown. Within the scope of this study it is
possible only to suggest some of the themes that have emerged in what
one could call linguistic nationalism.
Broadly speaking, modem nationalism is a phenomenon predomi-
nantly of the nineteenth century. Before that time there existed a var¬
iety of competing loyalties ranging from feudal allegiance, membership
of an estate, clan or guild, loyalty to village or town versus country,
faith or region. Some of these might be expressed in particular linguis¬
tic ways, such as the widespread use of dialectal forms or the use of a
certain language (e.g. French) among members of other groups (e.g.
the Polish, German or Russian nobility) or for specific purposes (e.g.
Latin in the liturgy of the Catholic Church). It was seldom the case,
however, that the preference for some linguistic forms stood in the way
of wider loyalties.
The idea of nation and the nation state became a central force only
after the French Revolution (see section 10.1.3). However, whilst lan¬
guage assumed a new importance in the identification with the state
and as a symbol of nationality, nationalism created a range of points of
linguistic tension. For example, nineteenth-century France was a highly
centralized state, and the only language which it was possible to speak
in the transaction of public affairs was French; this meant that the other
languages of France were excluded from the education system (suc¬
cessive French governments to the present day have defended and
maintained this decision). Such problems were even more acute in
larger territorial units such as the Habsburg Empire, where language
policy became a means of suppressing particular ethnic groupings. At
the same time, broader language-related trends (e.g. the Pan-Slav and
Pan-^vlordic movements) were used to weld common linguistic groups
together beyond national and imperial boundaries.
In all these cases (and in smaller countries, too) a new sense of
linguistic identity was asserted and, to some extent, created anew. Na¬
tionalist movements sprang up in many places, with sentiments that
were expressed in a variety of ways: in the use of dialect for literary
and other written purposes, in renewed interest in national myths, tradi¬
tions, legends and folklore, in efforts to standardize minority
languages, or in attempts to nurture a particular language variety. Feel¬
ings of nationalism were also expressed by more militant, even violent,
attitudes with which the more active members of minorities confronted
the majority, who more often than not sought to suppress them. In
effect, then, the linking of nationality with language found acceptance
in many parts of Europe, both at the higher ideological level and on
more practical and political grounds.
While language (or, more precisely, certain national languages) as¬
sumed a new importance, nationalism itself was a source of conflict.
A HISTORICAL PERSPECTIVE 229
Many countries experienced a number of developments (industrializa¬

tion, urbanization and, above all, centralization of the administration)
that invariably brought suppression of regional differences. There were
also problems arising out of the changes in the political map of Europe
which were the consequence of the emergence of new nation states
after the dissolution of the Ottoman Empire: some nations became in¬
dependent, others were reduced to the status of a minority within a
state or remained as they were before, a minority within another na¬
tion. By the end of the nineteenth century some forty-five million
people were members of linguistic minorities in various parts of
Europe (Stephens 1976).
The origins of the First World War were manifold, but nationalism
at both state and minority levels was a contributory cause. One of the
slogans of the day that appealed most to the inhabitants of many parts
of Europe was ‘For the rights of small nations’. The post-war settle¬
ment in Europe can be seen as an example of the problems that can
result from the attempt to reconcile linguistically-based national feel¬
ings with economic and territorial realities. The point of departure was
the principle of self-determination contained in President Wilson’s
famous Fourteen Points for Peace, proclaimed in 1918 before the
American Congress just prior to the official end of the war. At the
same time, the dissolution of the Russian Empire resulted in the setting
up of the new independent states of Estonia, Latvia, Lithuania and Fin¬
land. Poland, Europe’s whipping boy, who in her long history had
endured so many changes to her borders, on this occasion regained
some territory and with the establishment of the Polish Corridor ob¬
tained access to the Baltic Sea. Most of the states that emerged after
the demise of the Austrio-Hungarian and Ottoman Empires were faced
with intractable problems from the outset, either because sizeable mi¬
norities were included within them (for instance, up to 40 per cent of
the populations of Poland and Czechoslovakia consisted of minorities
that had little in common with each other) or because their minorities
made territorial claims against each other.
The emergence of fascism in Spain, Italy and Germany was to have
a profound effect on linguistic minorities. Language was considered an
essential part of nationality and a symbol of national identity and al¬
legiance. At the end of the 1930s the Basques and the Catalans, in
Spain, faced another kind of difficulty. During the Spanish Civil War
they had fought ‘on the wrong side’, that of the losing Second Repub¬
lic, which had given them a considerable degree of autonomy during
its short existence, and therefore* they found themselves in 1939 at the
mercy of the victorious fascists. Franco’s dictatorship retaliated with a
most severe repression of the so-called ‘traitor provinces’, which was
to have serious consequences for the survival of the Catalan and Bas¬
que languages.
The Nazis showed total disregard for the rights of non-German mi¬
norities in Germany and they considered all German-speaking
minorities outside the national territory as belonging, or wanting to be¬
long, to the Reich. Their justification of expansionist policies was that
they were efforts to bring again all Germans to where they belonged;
heim ins Reich, ‘home to the Reich’, was the slogan used. The annexa¬
tion of Austria and the Sudetenland (the German-speaking part of
Czechoslovakia), as well as the invasion of Poland, France, Luxem¬
bourg and Belgium, were presented in this light. Not surprisingly,
many minorities did not want to be identified with Nazi Germany, just
as many had failed to see themselves as having much in common with
the Kaiser’s First Reich. Alsace, Luxembourg and Altbelgien had, in
the years since the end of the First World War, developed their own
sense of national identity, in which dialect, cultural heritage and shared
history were the most important ingredients. While they used standard
German as the preferred form for written communication, their allegi¬
ance was towards the French, Luxembourgish and Belgian states. An
event from the Second World War serves to illustrate this. In 1940 the
Gerqians invaded Luxembourg; in order to demonstrate to the world
that they were really just enabling the local inhabitants to return to the
Reich, they held a referendum in which they asked them whether their
native language was German or French: over 90 per cent of those
questioned answered neither the one nor the other, but wrote ‘Letze-
buerguesch’ on the ballot papers. The cases of Alsace, Belgium and
Luxembourg show that linguistic minorities can achieve a sense of
identity within some countries and that problems need not be insoluble.
At the end of the Second World War Europe’s minorities had been
reduced considerably: in 1945 their number was estimated to be about
ten million, one-third of what it had been before the war (Stephens
1976). The excesses of fascist regimes, dejxirtation, resettlement, im¬
migration and the ravages of war had all contributed to this decline.
Neither the Allied powers nor the United Nations had anything con-
stmctive to offer to the European minorities, as there were no
provisions in international law that could be called upon in cases of
conflicts breaking out over disputed areas involving linguistic mi-
THE CONTEMPORARY SITUATION IN WESTERN EUROPE 231
non ties. The Turkish invasion of Cyprus in 1974 was a vivid reminder
of this.
11.5 The contemporary situation in western Europe
The survival and protection of the culture and language of linguistic

minorities therefore depend largely on their own efforts and on the
policies of the host governments towards them. There is only one inter¬
national body which campaigns on behalf of minorities in Europe, the
Federal Union of European Nationalities (FUEN). In its statutes a na¬
tionality is defined as ‘a national group which manifests itself by
criteria such as its own language, culture or traditions and which con¬
stitutes in its native soil no proper State or is domiciled outside the
State of its nationality’. Although the FUEN is an independent body, it
does act as a pressure group as well as a centre of information, suppor¬
ting and advising institutions such as the Council of Europe and the
United Nations on matters concerning minorities. For instance, for
years it has been campaigning for international legislation to cover the
rights of minorities. Its aim is to introduce into international legislation
the principle that a member of a linguistic minority has the right to use
his/her own language in education, administration, court cases, etc.
A step forward was taken by the European Community in its 1977
Directive, which set out to establish the principle that all children
should receive teaching in their mother tongue (see Appendix A). Al¬
though it was aimed at the children of migrant workers in particular, it
was thought that educational provision for minorities in general would
benefit. However, progress on the whole has been very slow. In
general, it can be said that those minorities who find themselves mem¬
bers of a federal state or of one that grants some degree of autonomy
to its regions fare better than those who have to live in a centrally
administered system. In countries such as Britain and France the con¬
flict between central government and linguistic minorities has
traditionally been bitter. Centralist policies aimed at suppressing ex¬
pressions of the minority’s national identity, such as their language and
culture, have brought about a serious threat to their very survival. In
the course of its short existence Belgium has been torn by many con¬
flicts which have caused the downfall of many governments. The
causes are generally perceived as being linguistic, although (as is
usually the case in disputes involving minorities) a host of other factors
are also involved. Until 1960 Btelgium was a highly centralized state,
but since then the Belgian Constitution has been changed several times
to allow for a greater degree of federalism. Decentralization was seen
as the only answer to solving the country’s problems brought about by
the mutual antagonism of the Flemings and the Walloons. Switzerland
has assumed the role of Europe’s most successful multilingual state.
There is no doubt that relative linguistic harmony has been achieved in
the Swiss Federation because the four major language areas enjoy vir¬
tually full political and cultural autonomy.
Other countries in western Europe have also seen a conversion to
federalism or regional autonomy. In Italy the process of devolution of
central power was set in motion in 1947, and it has taken over forty
years to be finalized. But Italy is not a fully federal state. Although the
regions have wide-ranging powers in some areas, the central govern¬
ment in Rome retains considerable power in matters of finance. Also,
Italy’s regions differ greatly from each other in terms of politics, cul¬
ture and economic prosperity; so long as this imbalance persists,
regional and minority problems are bound to remain. Thus unrest in
South Tyrol has flared up periodically ever since it became part of Italy
under the Treaty of Versailles - there was a spate of bombings and
angry demonstrations as recently as 1986.
Spain is Europe’s latest convert to something approaching federal¬
ism. Franco’s death in 1975 and the country’s return to democracy put
an end to over forty years of centralist government. The new 1978
Constitution established principles under which a number of regions
achieved a considerable degree of autonomy in the years to follow,
amohg them the linguistic minority areas of Galicia, Catalonia and the
Basque Country. In all three measures have been initiated to strengthen
the local language - a daunting task in view of the problems posed by
large numbers of native speakers of Castilian Spanish who have settled
in the regions concerned and the fact that the autonomous local gov¬
ernments receive no funds from the central administration specifically
allocated for their endeavours of language recovery (although the re¬
gions can decide how they spend their own funds).
The best guarantee, or at least chance, for the protection of mi¬
norities, then, seems to lie in some form of federalism, at both national
and international levels. Ever since the idea of European integration
was first discussed, federalists, such as the Breton leader Yann Four
with his notion of the ‘Europe of a hundred flags’ and the former Ger¬
man Chancellor Helmut Schmidt with his demand for a polycentric
USE AND MAINTENANCE OF MINORITY LANGUAGES 233
Europe, a ‘Europe of the regions’, have played an important role in the

promotion of European unity. However, such suggestions have been
vigorously opposed by supporters of other ideas, e.g. functionalism.
France, one of the largest and most centrally situated states of the Eu¬
ropean Community, is also its most centrally administered member
country. Successive French governments have insisted that a united Eu¬
rope should be a ‘Europe of the States’, and this view has found favour
with Britain, where federalism has never been seriously considered.
Where does this leave Europe’s minorities? Those who enjoy some
form of autonomous or federal status obviously have greater possi¬
bilities to preserve and to foster their national identity, culture and
language than those who receive no support from a centralized govern¬
ment. The only agency through which the latter can work towards an
improvement in their situation is the Bureau for Unrepresented Nations
in Brussels. Indirect help and encouragement (for example, in the form
of study scholarships and grants for research into matters of interest to
minorities) is provided by such institutions as the Council of Europe
and the European Commission. But of course such measures are a far
cry from anything that might enable any minority the form of self-
determination that it might seek.
11.6 Use and maintenance of minority languages
The extent to which a minority is able to use and maintain its language
depends on the inter-relationship of a large number of political, econ¬
omic and social factors. In no two minorities is this interplay of forces
identical. Minorities vary in size, geographical situation, social compo¬
sition and economic strength, and the political status that they enjoy
may range from almost full autonomy to total suppression. Yet there
are also some features which many of them share and which (either
alone or in combination with other factors) have posed a threat to the
survival of the minority’s language. For centuries minorities have suf¬
fered various forms of political persecution. Imposing restrictions on
their rights (e.g. to own land or to vote), banning the use of the mi¬
nority language in schools, courts and local administration and
generally neglecting local cultural traditions are all negative measures
that have been taken by centralist governments. In eastern Europe
many minorities have even, in the course of history, become virtually
extinct as a result of such policies coupled with large-scale expulsions
or extermination and subsequent repopulation with peoples of different

ethnic stock; and their languages (for instance, Yiddish, Latvian, Li¬
thuanian and Estonian, and German in the USSR) have suffered with
them. Another, perhaps less violent form of political suppression, is
total denial of the existence of a minority. This is, for example, what
the Turkish government does with respect to the Kurds who live in the
south-eastern comer of Turkey. The Kurds have their own language,
culture and religion, and many of them do not speak Turkish, which
means that government officials and members of the armed forces who
are sent to the region to fight Kurdish separatist guerrillas have to learn
their language - yet Ankara refuses to allow them minority status.
Most minorities have suffered from a common economic fate,
namely mral decline and depopulation. This has happened in Wales,
Scotland and Ireland, in Galicia and in practically all the minority
areas in France, in Friesland as well as in the Romansch-speaking part
of Switzerland, in both Corsica and Sardinia. Sometimes the rural exo¬
dus was speeded up by industrial exploitation in other parts of the
minority area, as happened with coalmining in Wales, Wallonia, Al¬
sace-Lorraine and Sweden. Then, after a period of over-production, the
coal-mining industry all over Europe slumped into recession, and these
areas became subject to further decline and emigration. More prosper¬
ous regions, such as the Basque Country, Catalonia and some parts of
northern Italy, experienced large-scale immigration as their industries
expanded, and with the influx of native speakers of the majority lan¬
guage the homogeneity of the region and communication in the
minority language became diluted. As regards the development of tour¬
ism, in many minority areas it has brought mixed blessings. It can
boost the local economy and provide jobs for the population, but it also
means that for at least part of the year large numbers of non-minority
language speakers stay in the area, and so communication in the ma¬
jority language becomes a necessity, thus putting further pressure on
the minority population, as happens in many areas along the Mediter¬
ranean coast of Europe and in the Alps.
There are also social factors that may affect language maintenance
negatively among some minorities. One of these is conscription. There
is evidence of this, for example, in Lepschy and Lepschy (1977), who
discuss the development of standard Italian and the decline of regional
languages and dialects in Italy. These authors point out the levelling
effect that conscription has had, and is still having, in this country. The
same is true in France, Spain and a number of other European nations
ASSIMILATION OF MINORITIES AND CULTURAL PLURALISM 235
with a tradition of conscription. For a period of time, young men from

different parts of the country have to communicate using the majority
language, and this interrupts or brings to an end their use of the mother
tongue and weakens its position.
11.7 Assimilation of minorities and cultural pluralism
So far we have discussed some of the political, economic and social

factors that in the course of their history have threatened the survival
of minorities. Ultimately, whether a minority is able to maintain its
cultural and linguistic identity will depend on the policies of the state
of which it forms part.There are two opposing stances that a state can
decide to take vis-d-vis its minorities: either it adopts a policy of cultu¬
ral pluralism (which will normally go hand in hand with some form of
federal organization) or it insists on the assimilation of the minorities.
Edwards (1985) notes that assimilation has many connotations, two
of which he describes as conformity and complete homogenization.
The former implies that groups adapt themselves to the dominant
mainstream culture which itself remains stable. Within the English-
speaking world this kind of assimilation has also been termed ‘Anglo-
conformity’, a reflection of the preferred policies in the English and
American colonizing contexts. The idea of the ‘melting pot’, although
used mainly in relation to the United States, more properly reflects the
idea of complete homogenization, where all elements of society, in¬
digenous as well as immigrant, minorities as much as the majority
group, intermingle to produce a new identity for all. The development
of post-colonial societies in many central and south American countries
probably corresponds more closely to the ‘melting pot’ image than the
case of their northern neighbours; Both types of assimilation have been
criticized as ‘a surrender of identities before some New-World jugger¬
naut’ (Edwards 1985: 104).
In some national contexts the assimilation of minorities has pro¬
ceeded along more moderate lines and in different ways, resulting in
the partial retention of some ethnic characteristics (e.g. religion or lan¬
guage) while assimilating along other cultural lines, such as giving up
certain customs, dress or language(s).
There are many features that make up the ethnic and cultural dis¬
tinctiveness of a linguistic minority, such as social institutions (e.g.
family structure and common law), religion, social customs and cultu-
ral traditions (e.g. food, dress, feast days, marriage, birth and death
rituals, etc.). A comparison of minorities which have been partially as¬
similated into a majority society is fascinating as it shows that certain
ethnic features are given up more readily by some than by others. It
would be difficult to prove that any one characteristic is more import¬
ant than others for the survival of the minority, but religion and
language are frequently found among the central features. Thus for the
large number of Jews who left their native Russia, where they had
lived for a long time in Yiddish-speaking communities, at the begin¬
ning of the century to settle in Britain or the United States, the
maintenance of religion (which, it is true, embraces many social and
cultural aspects of Jewish life) took precedence over everything else,
including the language. The same can be said of many second- and
third-generation immigrants in Britain, who may have succeeded in
preserving their religion but perhaps not their language, e.g. the Greek
and Turkish Cypriots, the Armenians and the Italians.
Those states which have pursued policies geared towards pluralism
have provided their minorities with much better opportunities to de¬
velop. In present-day Europe there are not many nations that can claim
to be genuinely pluralistic (Switzerland, which has a long polyethnic
and pluralist tradition, comes close to one, at least with regard to its
indigenous minorities). Integration results, in such states, when positive
relations with the mainstream society are encouraged by members of
b(^h the minority and the majority groups. Edwards (1985) refers to
this as voluntary pluralism. If, on the other hand, this does not happen,
segregation (either self-imposed or forced) is the result. Clearly, some
form of pluralism or modified assimilation is preferable as an approach
towards minorities, as either makes it possible for them to maintain
part of their identity and at the same time allows for a normal working
relationship between the groups and with the institutions of the state.
Nevertheless, there are reports from all parts of the world about ethnic
conflicts - the ideal of harmonious diversity, even in apparently plu¬
ralistic societies, remains in most cases unattainable. A quite
considerable degree of consensus is necessary to keep ethnic harmony,
but this consensus can be achieved only if all sides are prepared to
give up certain demands.
SEPARATISM AND SEPARATE IDENTITY 237
11.8 Separatism and separate identity
While many young states in Africa and Asia are still in the process of
nation-building, which requires a great deal of effort in order to over¬
come the problems of ethnic diversity, the old world (including the
United States) is experiencing a trend which has been termed the eth¬
nic revival. Whether we are faced with a genuine revival (in the sense
of renewed interest in, and resurgence of, things ethnic) and whether or
not ethnic minorities are now in a better position to express their ident¬
ity and pursue their claims are questions that fall beyond our remit. It
is clear that ethnicity is in people’s minds in many European countries
in one form or another. It is also a fact that certain well-organized
separatist movements have attracted widespread support for a range of
activities (both legal, militant and of an extreme kind) from members
of minorities and even some others: lobbying members of parliament,
putting up candidates for local government elections, taking part in
demonstrations in the regions and in the capital, pursuing legal cases to
the highest level when minority issues were involved (perhaps under
the protection of European legislation where neglect by a national gov¬
ernment was involved) - as well as the more fanatical actions resorted
to by separatists, such as daubing road-signs, putting up signs in the
regional language and, more exceptionally, going on hunger strike and
carrying out bomb and arson attacks or kidnappings. Wales, the Bas¬
que Country, Brittany, Corsica, the Jurassim and South Tyrol are all
places where separatist movements have been active in trying to
achieve their aims of gaining independence or some form of autonomy
from the majority government, in some cases more successfully than in
others. Their common belief is that if their activities and demands are
adequately reflected in the media, the attention of the public will be
drawn to the various issues relating to minorities. Providing that acti¬
vities are kept within the bounds of legality, this kind of publicity can
have a positive effect on general attitudes and foster greater ethnic and
linguistic tolerance and, in turn, benefit mutual relations. Several mi¬
nority groups have gained a larger degree of independence and
recognition in the past decade or two in this way, and there can be
little doubt that this has come about as a result of, among other things,
pressure exerted by separatists.
In discussing the basis of separatism, Williams (1984) points out
that the belief of nationalist leaders in their nations’ uniqueness is in¬
variably followed by a claim for independence. Sooner or later.
minority leaders come to see that their nation or ethnic group will not
be able to realize its full potential so long as it remains a constituent
part of a state. Smith (1982: 22) presents this idea quite clearly when
he says: ‘The watchwords of ethnic separatism are identity, authenticity
and diversity. . . It seeks through separation the restoration of a de¬
graded community to its rightful status and dignity, yet it also sees in
the status of a separate political existence the goal of that restoration
and the social embodiment of that dignity’. For a group of people to be
able to claim a separate identity it must show that it is different from
the majority in what Williams (1984) calls cultural infrastructure. It is
not enough for a group to identify with a particular territory where it is
concentrated. Through its shared history and sociocultural heritage the
group must have retained distinctive ethnic features which set it apart
from the mainstream society of a particular country. In Europe religion
and, particularly, language have always been the most powerful indica¬
tors of ethnic identity. Language is a real and demonstrable expression
of separateness, whereas ethnicity, strong sentiment though it may be,
is somewhat vaguer. Following Williams, we can argue that ethnic sep¬
aratism incorporates descent as the basis of individual and group status,
it provides spiritual confirmation, and it explains its uniqueness by a
shared history of suffering and isolation. Williams argues that ethnic
separateness sustains the myths of origin, development and uniqueness.
X.
11.9 Language and separateness
Language is the most powerful means of expressing separateness. On

an immediate functional level, as a means of communication, it serves
to include group members and to exclude those who are not, i.e. it acts
as a group barrier. On a broader cultural plane it provides a link with
the past, as well as continuity and prestige, particularly if the minority
has a rich literary tradition, as is the case with most minorities in
Europe. It is also able to unite different sections of the group (e.g.
different social classes), just as it can be instrumental in cultural divi¬
sion. It is undoubtedly the strongest barrier to assimilation and, not
surprisingly, it has played a key role in separatist movements. These
have always been most successful when they have been able to appeal
to a wider group by combining ethnic and regional-based grievances
with a claim for the preservation of religion or language — in other
words, when a coherent political organization has enabled them to rely
ATTRIBUTES OF MINORITY LANGUAGES 239
on the minority’s widely held sense of separateness and find appropri¬

ate expression for their perceived inequality vis-a-vis the state’s
distribution of economic resources. In their campaigns language has
always been used both as a means of communication and as a symbol
of group identity.
In an earlier study, Williams (1982) covers ethnic separatism in four
different geographical areas: Wales, Brittany, the Basque Country and
Quebec. He argues that in all cases language has been accorded group¬
defining significance and that the conflicts experienced by these
minorities (as well as others) can be described as language conflicts.
The primary concerns of these minorities may not be genuinely of a
linguistic nature, but since language is such a powerful divider of
groups and can be a barrier to communication, it also makes com¬
promise difficult. Williams (1982: 186) claims that ‘. . . in multi¬
lingual societies differentiation according to language is unavoidable. .
. Since coalitions rarely form across communication barriers, this pat¬
tern reinforces the language divide, predisposing both parties to
conflict rather than co-operation’. This view appears to offer little hope
for the peaceful coexistence of minorities and majorities, but it does
explain why separatism in our modem times is still an issue. It may
also serve to remind us that in order to ease what is termed language
conflict we have to find solutions that take into consideration both the
importance of language maintenance for minorities and the socio¬
economic and political grievances that minorities feel that they have.
11.10 Attributes of minority languages
So far the discussion has tried to bring together some of the sociocultu¬
ral factors that affect all minorities and the features that form part of
their identity. But probably the most fascinating aspect of the study of
minorities is the linguistic one. There is an extraordinarily varied pat¬
tern of language use to be detected: no two minorities will turn out to
have followed the same linguistic route or be going towards exactly
the same linguistic future. It is true for all minorities that language
maintenance is by no means guaranteed, but there are many forms of
(language) life before (language) death.
In a state where minority languages are not proscribed, linguistic
pluralism can take a variety of forms. A minority language can enjoy
the legal status of a national language, such as Irish Gaelic in Ireland,
or that of the official language of a particular region, as happens with

Flemish in Flanders and Catalan in Catalonia. It may be considered a
region’s language by the majority of its inhabitants without having any
particular legal status, as Ukrainian or Alsatian, or it may be one of
two officially recognized languages in a given area, e.g. Welsh in
Wales. It can also lack any kind of official status and still be employed
by a state’s authorities for specific purposes, as is the case with Span¬
ish in New Mexico (USA) or Urdu as used by some institutions in
Britain.
In the European context (perhaps more than elsewhere) the presence
or absence of official legal status can usually be taken as a good indi¬
cator of the state of the language, the prestige it and its speakers enjoy,
and its active use (how frequently and on what occasions they speak or
write it). As regards the linguistic state of the language, it is easier and
less controversial for the authorities to afford legal status to a minority
language (e.g. Catalan, Welsh, French in Canada) that has been stand¬
ardized and can be related to a body of written documents and books,
both of a literary and non-literary kind, than to a minority language
(such as an emerging Creole) that has no literary tradition or a lan¬
guage which has not been standardized (e.g. Romansch), or is
considered by the authorities to be a dialect (e.g. Alsatian see Chapter
12).
The prestige a language has depends on a number of factors relating
to the language itself, its users and the attitudes of the majority. If a
language has a standard and a literary tradition, and it is used by most
minority members, representing all sections of society, and for all
forms of daily communication, then it is likely to enjoy more prestige
both within the community and outside it than if it is only used for
certain purposes by some sections of the group. As so many minority
areas have suffered economic depression and loss of investment in new
technologies, their inhabitants have become associated with rural back¬
wardness and their language may have been tainted with the same kind
of stigma. It is this factor that provides the most challenging task for
language planners to overcome. For it is generally accepted that in
order to secure the survival of a minority language as a living entity it
is necessary not only to gain legal recognition for it, but also to prove
that it has at its disposal all the linguistic resources needed for success¬
ful communication in a modem, industrially advanced world.
Minorities which use a language that is official elsewhere (e.g. French,
German, Italian) clearly have an advantage over those who speak Ian-
ATTRIBUTES OF MINORITY LANGUAGES 241
guages that for centuries have been labelled dialects or vernaculars by

the majority in an attempt to devalue them and their speakers.
Linguistic pluralism in a minority area can be territorially based or
individually based, or we may find some combination of the two.
Multilingualism in Switzerland and Belgium is based on the principle
of territorial monolingualism, which means (as we saw in Chapter 8)
that only one language is the official one in a particular area; bilingual¬
ism will be a feature of individuals rather than the community as a
whole. However, in the Swiss Canton of Graubiinden virtually all in¬
habitants who speak Romansch are equally proficient in Swiss
German, i.e. the whole community is bilingual. In many of the Soviet
republics the local language (or one of several regional languages) is
used by the majority of the population for most of the time; yet Rus¬
sian, as well as being used by some members of the population all the
time, is the norm for certain functions such as higher education or ad¬
ministration. It is a common pattern in large, linguistically diverse
states, that the language of the ruling elite is employed for certain
high-ranking functions, whereas the regional language is used for all
other purposes by the majority of the population. But everybody is
supposed to know the language of the rulers, and since it is usually
favoured in and for education, the long-term aim appears to be to re¬
duce the use of the local language.
In the case of most of Europe’s minorities the number of monolin-
guals (in the minority language) is diminishing quickly, so that one can
probably say that all the members of the younger generations are either
bilingual or monolingual in the majority language. Whether the state
institutions (administrative, educational, welfare, etc.) in these minority
areas are (wholly or partly) bilingual depends on the legal status of the
minority language - in Wales some are, at least in certain areas, in
Brittany and the French Pyrenees they are not. Thus in any minority
area we find three types of speakers:
(1) Monolinguals in the ethnic minority language (usually the very

young and the older generations);
(2) Monolinguals in the national language of the state (numbers may
range from a minority of the population to the vast majority);
(3) Bilinguals.
11.11 Non-indigenous minorities in Europe
Non-indigenous minorities are those ethnic groups who are not long-
established members of a state. They comprise those people who have
moved to their current place of residence, or whose parents or grand¬
parents did so. These ‘new’ minorities (Churchill 1986) are the
protagonists of one of the most profound changes in the social struc¬
ture of many European countries in post-war years. The phenomenon
as such is not an entirely new one, as there have been many instances
in the history of the world of groups of people who changed their
domicile because they were forced to leave their native countries for
political or religious reasons (e.g. the Huguenots who in the sixteenth
and seventeenth centuries settled in Prussia, England and Denmark, or
the Hutterites who in 1874 emigrated to America) or because they
were encouraged by particular rulers to settle in their territories for
specific reasons (e.g. the Dutch who were invited to cultivate the mar¬
shes in Lincolnshire, England, in the seven teeth century). In most cases
these minorities eventually became assimilated to their new societies.
What is new about the present-day non-indigenous minorities is simply
the sheer numbers which are involved, both in terms of ethnic groups
and in terms of individuals.
Immigration (change of permanent residence to another country)
and migration (temporary residence away from home) are the two phe-
norpena that have created new minorities in Australia, the United
States, Canada and Europe. As immigration countries, the former three
encouraged large-scale immigration for nearly two centuries, and this
entailed a formal change of nationality and national allegiance on the
part of those involved. In Europe, France and Britain were the fa¬
voured places of destination for a long time, but up until the outbreak
of the Second World War immigration took place on a comparatively
modest scale and involved, in the main, people from other European
countries. Britain, for example, saw different waves of immigrants;
Germans who after the abortive 1848 Revolution had to flee their
country; Jews from Eastern Europe who fled from Tsarist and later
Bolshevik pogroms; Ukrainians who left their country after it at¬
tempted unsuccessfully to become independent in 1920; Poles in the
1920s and 1930s; and Jews from Germany and other European coun¬
tries who escaped Nazi persecution. The situation changed quite
dramatically after 1945 when France, Belgium, the Netherlands and
Britain received large numbers of immigrants from their former col-
NON-INDIGENOUS MINORITIES IN EUROPE 243
onies. Most of them came for economic reasons, and many (at least
initially) may have considered their stay as temporary. They tended to
settle in areas of industrial concentration and, in view of the numbers
involved, they were able to form their own communities. Until the Eu¬
ropean countries involved changed their nationality laws in an attempt
to curb the rise of immigration, those who came usually shared the
nationality of their former colonial masters, although they did not have
a common ethnic identity or (in many cases) even speak the same
home language. But the authorities somehow assumed that immigrants
would be able to communicate in the language of their new country,
and it was only with the realization that the children of these new
residents did not know the school language that a language problem
became acknowledged. It took some European countries a long time to
realize that the new immigrants would not assimilate quickly or easily.
The acceptance that the countries involved have become multi-ethnic
and multicultural, as well as multilingual, started only in the late 1970s
- the Netherlands proved more progressive in this respect, Britain and
France less so.
Migration of workers and their families started in the 1950s with the
expansion of the economies of western Europe. For three decades mi¬
gration moved in a south-to-north direction, involving millions of
people (in terms of numbers, it has been surpassed only by European
immigration to the United States). Most migrant workers came from
Turkey, Yugoslavia, Greece, Italy, Spain, Portugal and Morocco to
work in France, the Federal Republic of Germany, Switzerland, Den¬
mark, Sweden, Belgium, the Netherlands and, to a lesser extent,
Britain. At the beginning of the 1980s nearly twelve million people
were involved. Although most migrants settle in their new country of
residence on an ostensibly temporary basis, many of them become de
facto permanent residents, staying far beyond the initially intended four
or five years. In the case of, for example, the Turks in the German
Federal Republic, or Moroccans in the Netherlands and France, their
numbers are so large that they constitute sizeable communities, often
closely knit, with little need for outside contacts — for which the only
help needed is the services of interpreters, i.e. bilingual minority mem¬
bers. This, together with the intended temporary nature of their stay,
the work and legal restrictions imposed on them, and a generally nega¬
tive attitude on the part of mainstream society (both at an individual
and institutional level) can make their integration more difficult than
that of immigrant communities.
A number of reasons can be suggested to explain the fact that these

minorities pose particularly complex problems with regard to the lin¬
guistic situation and the host country’s appreciation of the issues
involved. One is that the majority of migrant workers come from the
most underprivileged sections of their home society, which means that
their standard of formal education can be quite low, often involving
illiteracy or a lack of familiarity with the standard language used in
their country of origin. A further reason is that, since communication
between them and the representatives of their countries (e.g. teachers
and interpreters) can be difficult, mother-tongue educational provision
for their children is often rendered ineffective. A third reason is that,
with so many different groups involved, making special educational
provision for them is a complex task — and this complexity is frequent¬
ly considered a convenient excuse for not tackling the problem at all. It
is the children of migrant workers who are likely to experience the
most profound cultural, linguistic and psychological problems, and we
are only just beginning to realize what is really involved in these is¬
sues. So long as small numbers of migrant workers stay temporarily in
another country, they can be disregarded, or simply considered as a
special feature of a particular area. But when they stay in large num¬
bers and subsequent generations maintain their cultural distance, we
can talk of a genuine minority. At the present time, many second-
generation migrants are experiencing marginality, a feeling of not
reall^y belonging either in the native country of their parents or in then-
own country of birth and residence. It remains to be seen how, and to
what extent, they eventually establish their own identity as ‘German
Turks’ or ‘French Moroccans’ or ‘Swiss Spaniards’ (see Chapter 14).
11.12 The study of linguistic minorities
Linguistic minorities constitute a relatively new subject of investiga¬

tion. Whereas older studies showed a preference for the description of
the history, the language and the sociocultural heritage (particularly the
literary tradition) of individual minorities, the emphasis has now
shifted towards a sociolinguistic and comparative approach. This can
be seen as a reflection of widespread interest in ethnic revival and
ethnic identity. There is also a growing awareness of the sociocultural
problems faced by minority members which, in some cases, has led to
a demand for language rights, notably bilingual education. In 1986 the
THE STUDY OF LINGUISTIC MINORITIES 245
Commission of the European Communities published a Report on Lin¬

guistic Minorities, which is an indication that in political circles, too,
minorities are being taken into consideration more and more. This Re¬
port contains little that, at the level of proposals for political action,
could actually help to preserve linguistic minorities, but it does provide
a systematic collection of data.^
The investigation of multilingualism is necessarily interdisciplinary,
i.e. two or more social sciences must contribute towards the elucidation
of particular issues from various angles. Sociology and education are
the two disciplines which have made the greatest advances in the field
of minority studies, both driven on by immediate social and political
concerns. But interdisciplinary investigation further requires the estab¬
lishment of a theoretical framework within which research can be
carried out. It also necessitates the use of different techniques, and it
ultimately demands some form of synthesis. For research to be inter¬
disciplinary (not just multidisciplinary) there must also be integration
of concepts. In the context of linguistic minorities this integration can
become particularly problematic, as there are many variables at work
relating to the languages under investigation, the social groups who
speak them and their historical background. It is a challenging task to
have to take note of all the relevant factors. While, for example, it may
be relatively straightforward to describe and compare language use
among two given minorities, it may prove impossible to account for
the fact that one has largely lost its native tongue although the other
has not. One of the reasons for this difficulty is that no long-term
studies have ever been carried out in this area, since it is impossible to
undertake academic projects that span several generations. Other
studies (e.g. Edwards 1984; 1985; and Williams 1984) have pointed
towards the researchers’ tendency to project their own biases when dis¬
cussing minority issues. Linguistic minorities - whether indigenous,
immigrant or migrant - present western European societies with chal¬
lenges and questions that they have never before had to face, and it is
virtually impossible to remain dispassionately objective. Also, the so¬
ciolinguist’s and psycholinguist’s old problem of the ‘intimate
relationship between the observer and the observed’ (Gardner 1965)
may stand in the way of objective investigation. Not all researchers
appear to be aware of the fact that they view reality from their own
moral or ideological standpoint, and that this may influence their re¬
sults. It is therefore necessary for those interested in such research to
be wary of possible bias.
Another problem that the researcher of linguistic minorities faces is

concerned with access to reliable material. Short of carrying out one’s
own fieldwork, one has to rely on what other investigators have found
out from official data. With regard to the latter, national and federal
government agencies often have surprisingly few (if any) reliable
figures about speakers or users of languages other than the national
one. There are many possible reasons for this. Apart from the inherent
difficulties involved in large-scale data collection, there is the addi¬
tional problem posed by the fact that many countries have in the past
seldom carried out reliable censuses - even today they still engage in
this activity only every five to ten years. In any case, censuses rarely
incorporate questions relating to language use, although they may in¬
clude items about nationality - by which more often than not what is
meant is state (rather than ethnic) nationality, i.e. ‘British’ rather than
‘Welsh’, for example. In addition, a question about mother tongue can
be quite inconsequential if it is not backed up by further requests for
information about the frequency and function of language use. One
notable exception to this situation is Catalonia, which, after achieving
autonomy and adopting a policy of promoting Catalan with a view to
making it the real language of the region, included in its 1986 census,
for the first time, several quite detailed questions about people’s use of,
and proficiency in, Catalan and Castilian Spanish. In this instance, the
information obtained will probably enable Catalonia’s language plan¬
ner^ to carry out their policies more efficiently. But census results can
also prove problematic, as has been demonstrated by the case of Bel¬
gium. Until 1962 the census questions on language (asked every ten
years) decided whether or not the boundaries of minority areas had to
be adjusted and whether provision for language services would be
maintained or withdrawn. But it became clear that the answers were
not always a truthful reflection of the prevailing linguistic situation, so
the aims of the census (i.e. to provide a sound basis for a fair distribu¬
tion of territory and facilities for the minorities involved) could not be
achieved. In the context of major linguistic reforms in Belgium in the
1960s, the linguistic borders were made definitive and the language
questions were withdrawn from subsequent censuses.
A further problem area is raised by the large number of illegal im¬
migrants and migrants who, not being entitled to live in the country, do
not figure in the official data. In western Europe their number is esti¬
mated (Power 1984) to be millions rather than thousands. Because of
their status, they do not enjoy any foiTn of protection or special provi-
NOTES 247
sion. They, and especially their children, represent a particularly

problematic section of the new minorities.
Notes
1. Studies on the subject of linguistic minorities have been under¬
taken, for instance, by Stephens (1976); Allardt (1979); Foster
(1980); Edwards (1985); Churchill (1986); and Hinderling
(1986).
Chapter 12
Case Study I: The Alsatians
12.1 The language
‘Of all the regions within the borders of the French state, Alsace has
one of the most clearly defined personalities’ (Stephens 1976: 341).
This statement may sound surprising in view of the extraordinarily
complex history of the region, which has been profoundly influenced
by two major nations (France and Germany). The Alsatians have gone
through many periods of conflicting loyalties, which have caused some
writers on the subject to talk about a ‘crisis of identity’ (e.g. Haug
1984) or ‘le particularism alsacien’ (Philipps 1975).
Apart from the approximately 200,0(X) speakers of Flemish in
French Flanders (which has a population of four million), the Alsatians
are the only long-established minority in France whose language is
closely related to the official language of a neighbouring state. They
also constitute the only minority in France whose speakers, some 1.3
million, are demographically concentrated: some three-quarters of the
region’s population are speakers of Alsatian.
Alsatian is a variety of German. It is similar to the Swabian and
Swiss dialects spoken to the east and south of Alsace, and to the Rhen¬
ish dialects of the north. Although it is sometimes used in a written
form (e.g. in folk literature, songs, poems, advertisements, captions in
newspapers), the written standard has traditionally been ‘Hochdeutsch’
(Standard German). Alsatian is thus basically a spoken language. Its
special features are particularly noticeable in the lexis, where a large
number of words and expressions are borrowed from French. Similarly,
the Alsatian variety of French (i.e. the French spoken by many Alsat¬
ians) is marked by German influence (e.g. intonation that shows the
transposition of the German tonic accent on the first syllable as op¬
posed to the last syllable in each group as in Standard French) and by
‘Alsatianisms’ in lexis and expressions.
THE LANGUAGE 249
MOSELLE - LORRAINE
BAS-RHIN ) ALSACE
HAUT-RHIN f
__ Present linguistic frontier

French-Alsatian
-Border of department
SWITZERLAND
Luccllc
Lulzcl
Alsace Lorraine
250 CASE STUDY I: THE ALSATIANS
Alsatian in fact consists of different dialect varieties. The four prin¬

cipal ones are:
(1) ‘francique mosellan’ (Moselfrdnkisch), spoken around the fron¬

tier with Luxembourg, e.g. Thionville;
(2) ‘francique rhenan’ (Rheinfrdnkisch), spoken in Lorraine, Alsace
Bossue, Wissembourg;
(3) ‘has alemanique’ (Niederallemannisch), spoken in the greater
part of Alsace;
(4) ‘haut alemanique’ (Hochallemannisch), spoken near the frontier
with Switzerland.
The Alsatian dialects thus range from Franconian, a central German

one, to Alemannic, which belongs to the Upper German dialect group.
These dialects are spoken as their mother tongues by the majority of
speakers throughout the region, although in the north this is true to a
lesser extent than elsewhere: in northern Alsace, as in the neighbouring
but quite separate region of Lorraine, the language most widely spoken
is a local variety of French. The absence of one single ‘standard’ var¬
iety of Alsatian is sometimes mentioned as a reason why it would be
difficult to teach Alsatian in schools. Lack of an agreed standard is not
an unknown problem in minority areas: for instance, in the Romansch-
speaking areas of Switzerland, or in the Gaeltacht in the Irish Repub¬
lic, speakers see this as a drawback, too. Gardner-Chloros (1983: 35)
make> the point that, unlike German dialect speakers, ‘Alsatians can¬
not, or do not wish to, approximate more or less to Standard German’.
One could add that, if this is so, this feature would distinguish them
from speakers of the various Swiss German dialects, who will switch
relatively easily from dialect to Standard German. In contrast, the Al¬
satians will usually switch to the standard variety of an entirely
different language, namely French (for instance, when talking to a for¬
eigner).
12.2 Historical overview
Geographically, Alsace (the ‘departments’ of Haut-Rhin and Bas-Rhin)

is shaped like a corridor in the Rhine Valley between north and south.
It is separated from Germany by the River Rhine and from the west of
France by the Vosges mountains. It therefore forms a distinct, self-con¬
tained area. As we know, political frontiers rarely coincide with
linguistic ones, but the latter are found more often along mountain
ranges than alongside rivers. This may explain, at least in part, the
HISTORICAL OVERVIEW 251
survival and widespread use of a distinct German dialect within the

French state. It has been observed that in this region more similarities
are found between the dialects spoken along a given horizontal line on
the French-German border (i.e. across the river) than between the var¬
ieties of Alsatian used in the north, near the frontier with Luxembourg,
on the one hand, and those spoken in the south, near the Swiss border,
on the other.
Given its geographical situation and topography, Alsace has been
exposed to innumerable invasions from the times of the Franks and
Alemans onwards. And ever since Charlemagne’s empire was divided
among his successors it has been claimed by the powers on either side
of the Rhine. However, because of its geographical situation, religion,
art, economy, language and culture, Alsace was always part of the
Rhine area (Rheingebiet), and it was therefore intimately linked with
the German world. Alsace’s more recent history can be divided into six
periods:
(1) Until 1648 Alsace was part of the (German) Holy Roman Empire.
In the Middle Ages, towns such as Strasbourg and Colmar were great
centres of literary and other cultural activities. Later they played an
important role in spreading the use of German as an official language
(when Latin was abandoned) and as a language used in church. The
first bible translated into German was published in Strasbourg in 1466,
well before Luther published his translation (1521).
(2) 1648-1870 Under the Treaty of Westphalia, at the end of the Thirty
Years War, Alsace was annexed to France. The rich economic and
intellectual life of the region was disturbed both as a consequence of
the war and as a result of Louis XIV’s policies. But Alsace s language
(initially, at least) was respected. Indeed, the region retained more in¬
dependence than other minority areas in France: for example, its
schools were not subject to interference, and Alsatians were exempt
from military service - two factors which contributed to the survival of
Alsatian as a language. The process of assimilation to France began as
soon as the aristocracy, the upper classes and the intellectuals were
forced to use French in their contacts with the new administration. But
in most areas of life in local society Alsatian remained the language
used for daily interaction, while documents and official papers were
written in German.
Stephens (1976: 343) quotes the Alsatian Andre Ulrich, who in

1789, just before the French Revolution, was able to declare: ‘There
are three hundred inhabitants of Alsace who do not know French for
every one who does’. Alsace’s identity as a region was dealt a heavy
blow by the French Revolution, which proclaimed the principle of ‘une
nation, une langue’ and set about enforcing linguistic uniformity, often
by strong-arm methods such as threatening to execute or deport those
unable to speak French. Apart from brute force, other factors con¬
tributed towards the spread of French during the nineteenth century,
reaching, as usual, the upper and middle classes first: French became
the language of secondary and higher education, and in 1853 it was
officially introduced as the language of primary schooling, although at
least one lesson of German was taught each day. The growing indus¬
trialization of the region, the development of transport and the
consequent increase of bureaucracy also favoured the spread of French.
But the Church remained a staunch supporter of Alsatian identity and
the German language. And although most educated people acquired a
knowledge of French, their native tongue remained Alsatian.
(3) 1870-1918 After the Franco-Prussian war Alsace changed hands

again. The process of germanization that was instigated by its new
masters was only made more palatable by the promise of greater re¬
gional autonomy. The late nineteenth century saw Alsace’s economy
flourish, and at the same time nationalist movements, both literary and
political, began to emerge. Many writers argued that the only solution
to the Alsatian predicament of being French in sentiment and German
in speech consisted in political autonomy for the entire region (that is
to say, in the institutionalization of Alsatian identity). This expression
of national identity should be viewed in the wider European context,
in which an interest in philology, nationalism (originally of a non-pol¬
itical nature), folklore and history during the nineteenth century led to
the development of many a national movement. Alsace’s dilemma, her
split national loyalties, came vividly to the fore when the Great War
broke out in 1914: 250,000 men from Alsace were conscripted into the
Kaiser’s army, while 20,000 volunteered to fight on the French side.
(4) 1919-40 Following the Treaty of Versailles, Alsace was returned to

France. As part of the highly centralized French state, it no longer
enjoyed any administrative autonomy, and Alsatian as a language was
firmly suppressed in all areas of public use. The late 1920s and 1930s
saw the emergence of political parties which, to varying degrees, cam¬
paigned for more administrative and cultural independence from Paris
and demanded the introduction of official bilingualism. With the rise
of Hitler in Germany, however, the Alsatian movement lost much of its
momentum. Nazism had the effect of furthering the French cause in
HISTORICAL OVERVIEW 253
Alsace, in the same way as it fostered distinctly anti-German senti¬

ments in neighbouring Luxembourg and Switzerland. In these two
countries this was to the benefit of the regional dialect (Letzeburgish
and Swiss German), which became upgraded in perceived prestige;
since the 1930s Standard German has been losing ground as a spoken
variety.
(5) 1940^5 During the Second World War Alsace was occupied by
German forces, who immediately proclaimed that Alsatians were Ger¬
man by origin (including those who spoke French!) and that German
was the language to be used. It was now the turn of French to be
banned from public life, and an intensive programme of germanization
was set in motion.
(6) At the end of the Second World War Alsace was once again re¬
turned to France, and again it became part of that country’s centralized
system of government, but this time no linguistic concessions were
made. German was dropped completely from public use, and it ceased
to be a language of instruction in schools. In fact, it was relegated to
the status of a ‘foreign language’. The justification given for these
measures was that they would enable French to ‘regagner le terrain
perdu’ (i.e. regain lost ground). The intention was to ensure, of course,
much more: that the region really became French, and this time for
good. And it has become French, to a large extent. Since 1945 French
has gained ground at every level of public life. Among members of the
older generations there are still those who have changed their nation¬
ality and official language two or three times in the course of their
lives, but everyone bom after the war has been educated entirely in
French. For the new generations, French is the language of all levels of
education, of administration, the courts of justice, the police, the media
- in short, all areas of public and intellectual life. Certain legal
measures have been taken, aimed at the preservation (but not active
encouragement) of regional languages in France. The passing of the
Loi Deixonne in 1951 represented a turning point in France’s linguistic
policy: it decreed that certain regional languages (Alsatian was not in¬
cluded) could be taught in their respective areas if a set of given
conditions were met. The Act was extended in 1961 to include Alsat¬
ian, but its provisions remained ineffective because of a long delay in
the publication of the necessary ‘decrets d’application’. A more suc¬
cessful reform came about in 1971 with the ‘Methode Holderith ,
which has resulted in more widespread teaching of German in second¬
ary schools, sometimes from the age of nine, usually for half an hour a
254 CASE STltDY 1: THE ALSATIANS
day, and in the encouragement of the use of dialect and the teaching of
dialect literature. Cellard (1976: 18) explains the measures as “pour
1 essentiel la methode Holderith consiste a mener parallelement, des
I’ecole primaire. I’apprentissage de I’aHemand a panir du dialecte. et
celui du traiK^'ais a partir d'une “potentialite bilingue” du petit dialec-
tophone’. Research on language use among young Alsatians indicates
that the bilingual potential remains domiant or. as the pessimists would
have it. is non-existent. Alsatian is not used at schcxd. and Standaal
German is taught as an optional subject, available in secondarv schcK^ls
only, tor a few hours a week. These factors, together with the ignor¬
ance ot parents and school staff on matters of bilingualism (which
often leads to the view that the acijuisition of the domimmt lamiuace
might be affected negatively) all contribute to the limited impact of the
Holderith refonns.
French is used throughout Alsace in public administration, the
media and education. All public notices, advertisements, RXid sisjns and
street names are in French. There are publications in Standaai German
(Alsatiiui is not used for written purpcises): but if newsfxrpers am in¬
volved they have, by law, to appear under a French title with the
subheading ‘edition bilingue* and a qiuirter of the material must be in
French. It can be taken as a subtle form ot linguistic manipulatit.>n that
articles addressed to young people and on sport must appx'ar in FmtKh;
they are allowed to be followed by a summary in Germrui. but that
seems, a rather meaningless concession (Gardner-Chloaxs 1983).
The Church is the only influential body that still uses .\l.satian. The
Catholic Church celebrates Mass in both French and German (or .\lsat-
iiui). but the Protestimt Churches use Standarxl German in their viuaous
liturgies. As in other minority areas, the Chumh. particularly at the
lower levels of the hieraivhy. is playing a significant arle in maintain¬
ing the minority liuiguage.
12.3 The use of Alsatian
Sociolinguistically. the language situation in .Alsace has been described

as diglossic. not in the ‘classical’ sense of Ferguson (1959). because
the two varieties involved are not fonns of the same language, but in
Fishman's (1967) extended version, w hich includes varieties of differ¬
ent languages in addition to the criterion of separation of function.
However, there are two factors that make Alsace's diglossia more
complex. One is the presence of High Gennan as the w ritten stmdaal
and the variety taught in schtx^ls (the use of a standaal v;irietv for
THE USE OF ALSATIAN 255
written purposes and a dialect one for oral communication is a com¬

mon enough phenomenon in dialect areas). The other factor relates to
the question of separation of function. One can say that, in general,
Alsatian is the language of the home, the family, interaction with close
friends and neighbours and small-scale commercial transactions, and
French is the language of public life and education. But this separation
of languages is not completely clear-cut. At one end of the scale there
are those Alsatians who are monolingual French speakers, and at the
other end there are those who use Alsatian in many kinds of communi¬
cative situations. For, whereas French can be used at all times and in
all situations (virtually all Alsatians can speak French), the use of Al¬
satian is determined by factors relating to the speaker, the listener, the
setting and the topic under discussion.
It is not unusual to encounter difficulties when trying to find re¬
liable data about language use among linguistic minorities. Most
authors on the use of Alsatian, for instance, quote data published by
the Institut national de la statistique et des etudes economiques
(INSEE). The latest census figures for language use were published as
long ago as 1962. According to them, 85 per cent of the population
spoke Alsatian and 80.7 per cent Erench. This represented an increase
of 14 per cent of French speakers since 1946. In 1980 the INSEE car¬
ried out a survey based on ten questions on the use of dialect in
Alsace, and the figures were published in 1981. They revealed that in
the region of Haut-Rhin 73 per cent of the population spoke dialect
and in Bas-Rhin 77 per cent. Ladin (1983) criticizes this survey for not
seeking information on such aspects as frequency of use of Alsatian,
the exact functions for which it is used, competence in the dialect and
young people’s expressed language loyalties. However, despite its
shortcomings (the most fundamental being that it was a survey and not
a census covering the whole population), the INSEE report does con¬
tain data that points towards specific factors which appear to determine
the amount of Alsatian spoken by a given individual. Age is probably
the most significant of these. Among the population at large the dis¬
tribution by age shows that, whereas 88 per cent of those over
sixty-five years of age claimed to speak Alsatian, the figure falls to 65
per cent for those between sixteen and thirty-four. Ladin (1983) carried
out a detailed study of the language use of Alsatian-speaking young
people (fourteen to sixteen years old). His questionnaire covered
seventy-eight items which were grouped into bundles and aimed at eli¬
citing information of the following kind:
(1) a sociolinguistic profile of the informants;

(2) language use in a number of communicative situations (fam-
ily/friends; inner monologue/dreams/thoughts; religion; public
life; cultural activities);
(3) the image of the three languages used in Alsace;
(4) linguistic self-evaluation;
(5) linguistic competence (consisting of a translation from French
into Alsatian and a task of labelling certain items in Alsatian).
Of the young people who took part, 77 per cent stated that they had
acquired Alsatian as their first language; for 23 per cent of them the
first language they had acquired was French. Only 4 per cent claimed
to have acquired the two languages simultaneously. When classified
according to parental background, the figures for use of dialect varied
considerably: whereas 97 per cent of children of farmers spoke Alsat¬
ian, only 64 per cent of those whose parents were clerical or
white-collar workers (‘Beamte und Angestellte’) did so. An even lower
proportion (44 per cent) of the children whose parents worked in the
professions or were top executives or high-ranking civil servants were
said to be speakers of Alsatian.
Another significant factor in the use of language is the distribution
by setting, i.e. urban versus rural.The French sociolinguist Andree Ta-
bouret-Keller carried out a survey between 1957 and 1962 that showed
a strong correlation between the use of Alsatian and the percentage of
the 'population working in agriculture, and she found farmers to be
extremely conservative in their linguistic habits. Gardner-Chloros
0983) points out, however, that ‘rural’ is not synonymous with ‘farm¬
ing population ; the latter now makes up only 4.7 per cent of the total
population of Alsace. But generally speaking it is true that more Alsat¬
ian is spoken in mral areas than in towns. The INSEE study covered
3000 households and found that 88 per cent of the inhabitants of rural
communities claimed to speak Alsatian, whereas in towns of between
10,000 and 15,000 people the figure was 69 per cent. We have already
seen that in the last century French was more widely used in urban
areas and Alsatian in the countryside, where the inhabitants’ contact
with French was much more limited than in the case of the bourgeois.
So there was then a clear correlation between setting and language use.
In modern times new factors have come into play which have been
highlighted by research. For example, Neville (1986) mentions the split
urban/rural as important, and she adds a new dimension, that of em¬
ployer/employee, quoting Gauthier (1982: 80): ‘Si, dans I’enterprise,
on parle encore en dialecte, c’est en franâis cependant qu’on s’adresse
THE USE OF ALSATIAN 257
au patron (en Alsace).’ (The use of ‘encore en dialecte’ is interesting as

it indirectly seems to point to the speaker’s perception of the future
demise of Alsatian.)
Naturally, the communicative situation is also relevant. Another sec¬
tion of the INSEE report, mentioned earlier, shows that of the families
questioned, 60 per cent said that they used Alsatian at home, 52 per
cent u.sed it in the shops and 30 per cent for social security transac¬
tions. When asked whether they spoke Alsatian at work, 71 per cent
answered that they used it ‘a lot’ or ‘quite a lot’, which seems to tie in
with Gauthier’s observations about language use among employers and
employees.
All these findings indicate that level of formality, rather than do¬
main/setting, is important: the more formal the situation the more
likely it is that French will be used. It should be remembered that in
minority areas such as Alsace switching between majority and minority
languages is the commonest phenomenon. In her study of code¬
switching among Strasbourg shoppers, Gardner-Chloros (1985) con¬
sidered four variables which she thought were influential in language
behaviour: (1) the type of shop/department store; (2) the sort of goods
purchased; (3) the age of both customers and salespersons; and (4)
whether they were involved in in-group or out-group conversations, i.e.
speaking among themselves or not. The pattern of switching altered
according to the different variables involved, but in most cases the re¬
searcher found that the switch was from French to Alsatian rather than
from Alsatian to French. The reason for this was difficult to establish,
but Gardner-Chloros speculates that the speakers’ uneasiness about
using French (because they were unaccustomed to it, or had difficulties
in talking in French over a sustained period of time) may have had
something to do with it. The study confirmed the impression of other
observers that French was generally perceived as being the prestige
form, required for use in more formal contexts, and that its use was
more frequent among younger than among older speakers. On the other
hand, the use of Alsatian was more prevalent in the downmarket stores,
in shops selling necessities rather than luxury goods, and in conversa¬
tion among older customers and staff; and it was also more frequent in
in-group conversations than in out-group ones.
It is also interesting to note that women claim to use all three lan¬
guages more often than men, whereas men tend to restrict themselves
to using just one or two languages. Women also said that they used
more French than men, particularly with their children once they had
.started school (Tabouret-Keller 1981). In the course of the research that
he carried out into attitudes towards the Alsatian dialect. Cole (1975)
found that when he asked Alsatians of various ages whether it was

important for them to speak Alsatian, between 83 per cent and 92 per
cent replied affirmatively; an even more significant finding was that, of
the 10-15 per cent who were just as happy to identify with ‘French’ as
with ‘French and Alsatian’, two-thirds were women. Ladin (1983), too,
found what he called a greater ‘Dialektfaulheit’ (i.e. reluctance to use
Alsatian) among mothers: 49 per cent of them used the dialect with
their children, as opposed to 57 per cent of fathers. In all the communi¬
cative situations that he studied, girls were found to use French more
consistently than boys. Why this should be so the researcher felt un¬
able to explain, but he pointed towards a possible connection between
women and, on the one hand, the jobs they often do, as many tend to
work in the service industries where French is required and, on the
other, their role as mothers, which entails contact with schools, tea¬
chers and administrative authorities. We know from classical
sociolinguistic studies, such as those by Labov in Martha’s Vineyard
(Labov 1963) and Tmdgill in Norwich (Trudgill 1972) that women
tend to be more conscious of, and interested in, the language forms
that carry overt prestige than men. The Alsatian data, although it is
sketchy, seems to confirm these findings.
Alsatian, therefore, should not generally be considered the language
of the home. The 1981 INSEE survey, which covered 1000 school-
children living in different parts of the region, showed that all spoke
French at home, albeit in varying proportions. In the centre of Stras¬
bourg only 8 per cent never or rarely spoke French at home, but in the
surrounding area the figure rose to 29 per cent, and in the countryside
72 per cent said that they never or seldom spoke French at home.
Comparing these figures with the ones about distribution by setting
(urban/rural), it seems clear that young people use more French than
their elders. This is confirmed by Ladin’s research (Ladin 1983), which
showed that, within the same family, the prime factor in the decision to
use French or Alsatian is age:
Language use within the family
parents to parents: 88 per cent Alsatian

parents to children: 53 per cent Alsatian
children to parents: 63 per cent Alsatian
children to siblings: 34 per cent Alsatian
42 per cent French
24 per cent Alsatian and French
THE FUTURE OF ALSATIAN 259
12.4 The future of Alsatian
The last set of figures given above is perhaps the clearest indication of
the decline of Alsatian. If the preferred language among the young
generation is French, and parents, especially mothers, use French rather
than Alsatian for their communication needs, and if, in addition,
schools do not encourage the teaching of Alsatian - how will the dia¬
lect be passed down to younger generations once the present-day
youngsters are older? From the most optimistic angle, renewed interest
in regional languages and dialects (which has manifested itself in
France as well as in many other parts of Europe) may become more
than just a quest for individual identity in an increasingly impersonal
and internationalized world, and it might result in more institutional
support for the Alsatian language and thereby enhance its prestige. But
so far there are no signs that this may happen. The Alsatians have been
profoundly affected by historical events and the resulting conflicts of
loyalties. More recently, Alsace has experienced economic recession
leading to a rapid decline in the area’s coal, steel, agricultural and tex¬
tile industries, as well as a reduction in investment by central
government. Unless Alsace is officially proclaimed bilingual and bi¬
lingualism is fostered by the state and local authorities, and unless the
area is able to determine its own economic and social destiny, there is
likely to be little improvement in the present situation of language de¬
cline. At the moment there is no strong nationalist movement
demanding autonomy and promoting the use of Alsatian, so the driving
force that could ensure the survival of Alsatian is missing. Indeed, all
the accounts of the present plight of Alsatian appear to indicate that, in
so far as Alsatians as a whole are aware that their language is in
danger, they feel no great sense of loss or urge to protect it: fewer than
half of those questioned by Ladin (1983) saw the use of Alsatian as an
essential ingredient in being Alsatian. Such attitudes may alleviate the
Alsatians’ dilemma of being tom between cultural and political loyal¬
ties towards France or Germany, and of being regarded as German by
their fellow Frenchmen while at the same time not wanting to give up
their allegiance to Alsace. As a linguist one cannot help feeling uneasy
about their lack of any sense of urgency to safeguard their language.
But, as has been argued before, language is only one of several group
markers. The demise of Alsatian does not necessarily imply that loss of
Alsatian identity will follow.
Chapter 13
Case Sudy II: The Catalans
13.1 Introduction: Catalonia and the Catalans
Catalonia is an area of special interest to the sociolinguist these days.

Although politically a part of Spain, it is a region that manifests its
cultural separateness clearly to the foreign visitor familiar with other
13.1 Linguistic minority areas in Spain

INTRODUCTION: CATALONIA AND THE CATALANS 261
— Linguistic frontier
-Main dialect division
-National frontier
-Regional frontier
13.2 The spread of Catalan dialects

262 CASE STUDY II: THE CATALANS
parts of Spain. The most striking feature is the use of its own language,
Catalan, both in the capital, Barcelona, and (to an even larger extent)
in the smaller cities and towns of the old Principality, in all walks of
life and for all purposes. As one travels through the area one becomes
gradually aware of the wealth of cultural and educational activities
going on in the language, supported by the authorities and by the pub¬
lic at large.
The sociolinguist attempting to describe and analyse the current lin¬
guistic situation in Catalonia will also find that it is as unique as many
Catalans claim. Every country is, of course, unique in its cultural and
social make-up, which includes the pattern of language use, particular¬
ly in countries where we find several minorities who maintain their
own language. Catalonia’s uniqueness lies in the fact that it does not
share many of the features that are common to most other European
minorities. Catalan society has a long history of stubborn resistance to
political and cultural assimilation, which central governments in Ma¬
drid have aimed at, and tried to enforce, often by the most draconian
measures. The result of the unattractiveness of central policies, coupled
with ineffective administration (particularly between 1812 and 1931),
has always been to foster the permanence of cultural and social ident¬
ity of Catalonia (the same can be said about another Spanish minority,
the Basques). The Catalan areas on the French side, on the other hand,
which were cut off from their political, economic and cultural capital,
hav^ assimilated French culture to a much greater extent, in response
to the administrative and levelling skills of a strong and prestigious
French state.
A comparison with other stateless nations in Europe that have also
kept a noticeable degree of linguistic and cultural differentiation, in
spite of prolonged suppression by the dominant majority, shows that
the Catalans manifest a much stronger degree of nationalism than
many other minorities. The Catalan sociologist, Salvador Giner, de¬
scribes the self-image of Catalans as being determined by their habits,
customs and cultural inclinations, and he mentions their proverbial
fondness for hard work, careful spending and profitable investment;
Taken together, their collective virtues are neither very heroic nor dra¬
matic. . . . They are precisely the virtues which have largely made
Western societies what they are today’ (Giner 1984). The Catalans’
open identification with these national traits is strong, and so is their
linguistic attachment. To be a Catalan means that you speak Catalan.
This prerequisite of language as an inseparable part of national identity
INTRODUCTION: CATALONIA AND THE CATALANS 263
does not always pertain among other minority nations in Europe, as we

have seen in the previous chapter. The combination of language and
perceived national character constitutes the ‘fet diferenciaT (differential
factor) or uniqueness of Catalonia. Allardt (1984) considers self-ascrip¬
tion to a minority group by its members as the basic criterion for the
existence of minorities. In the case of the Catalans, this self-categoriza¬
tion (as opposed to being categorized by others) has survived for
centuries, and it finds its expression today in the Statute of Autonomy
(1979), which declares that Catalan is Catalonia’s ‘own language’ {la
llengua propia) and in the Catalan government’s proclamation that it is
its ‘unquestionable right and duty’ to restore Catalan to ‘her rightful
place’. The tenet is that it is the right of every Catalan to speak Cata¬
lan. New residents, i.e. immigrants from other parts of Spain and their
descendants, must be enabled to learn it, just as those of Catalan origin
who have not learnt it yet, or not learnt to read and write in their native
tongue, have to be given a chance to recover their language.
Another aspect of Catalonia’s uniqueness is more academic. It lies
in the way in which the Catalan situation seems to defy, or at least
make more difficult, traditional categories. Basic concepts and terms
that have been adopted to describe language use and language patterns
in other linguistic communities often have to be revised, because they
turn out to be inadequate politically or emotionally charged, or both.
Different philological and linguistic traditions among Iberian and
Anglo-Saxon academics, and also the inconsistent use of linguistic and
sociological terminology by Spanish politicians, past and present, have
contributed to this. For over 250 years the central governments of
Spain sought to suppress the political, cultural and linguistic charac¬
teristics of Catalans, Basques and Galicians alike, and for the same
period of time these attempts were resisted. Terms such as ‘minority’,
‘language’, ‘dialect’ and ‘bilingualism’ were used by all sides involved
in the controversies, but with different meanings and connotations (see
also section 13.3).
In terms of their size and their economy, the Catalans are a signifi¬
cant minority within the European context. Catalonia is, together with
the Basque Country, Spain’s richest area in terms of per capita income,
and it is also the most advanced in terms of industrial and technologi¬
cal development; therefore it provides disproportionately high revenue
for the coffers of the Spanish state. This is not the position most other
European minorities find themselves in. Catalan history has a long and
distinguished tradition and is held in high esteem both inside and out-
1C
side the region. It has produced many of the most outstanding Spanish
artists.
It is also the only minority in Europe that has an influential govern¬
ment agency actively involved in extensive language planning (for
which it receives no central funds) with a view to establishing Catalan
as the ‘normal’ medium of communication within its territory (the
Catalans prefer to use the term ‘normal’, which to the outsider may
seem ambitious or ambiguous: it conveniently makes it unnecessary to
use ‘only language’ or ‘one of two languages’, the former being politi¬
cally explosive, the latter ideologically unacceptable; yet, at the same
time, the term clearly signals the commitment to language mainten¬
ance). These features are not shared by other European minorities, but
there are similarities in other areas: for example, most linguistic mi¬
norities have had to defend their language against the dominance of the
majority language and have been denied political independence. In¬
deed, in Catalonia, the centuries of oppression have taken their toll,
and the future viability of Catalan as the language of all Catalans is by
no means guaranteed, in spite of the emotional, ideological and finan¬
cial commitment of many people in the area.
13.2 Geographical and demographic overview
Catalonia is a politically autonomous region in the north-east of the

Iberian Peninsula. It comprises the four provinces of Barcelona, Tarr¬
agona, Lleida and Girona. Of the six million inhabitants (1989:
5,978,638, approximately 15.5 per cent of the total Spanish popula¬
tion), over half live in the metropolitan area of Barcelona which,
together with Tarragona, is the most industrialized part of the Princi¬
pality. Both provinces, but particularly Barcelona, have seen a very
heavy influx of immigrants from other parts of Spain, both before the
Spanish Civil War (1936-39) and, especially, during the period of in¬
dustrial recovery (the 1950s to the 1970s). Catalonia is surrounded by
other areas where Catalan is spoken: French Catalonia (the Roussillon
region) to the north of the border, Andorra, Valencia to the south, and
the Balearic Islands. In none of these does Catalan enjoy the same
political status or social prestige as in Catalonia itself, with the excep¬
tion of the tiny state of Andorra, where it is the official language.
Linguistic unity in ‘els Paisos Catalans’ has a long history and seems
to provide Catalonia with the justification to take on a leading role in
THE CATALAN LANGUAGE 265
fostering and strengthening linguistic and cultural links. Within Catalo¬

nia the majority of the population speak Catalan, although this
demographic concentration is distributed unevenly between urban and
rural areas. Outside the old Principality the speakers of Catalan are
geographically much more dispersed and socially more homogeneous -
both factors which tend weaken the maintenance of the language.
13.3 The Catalan language
Catalan is a Romance language. It developed in the later period of the

Roman Empire upon a pre-Roman substratum from a variety of Latin
spoken by soldiers, administrators, local peasants, sailors and crafts¬
men who lived to the north and east of the Pyrenees. It has been
described as a bridge language between the Ibero,-Romance and Gallo-
Romance languages, sharing many of the syntactic and morphological
features of the former and some of the phonetic characteristics of the
latter. Despite many similarities between Castilian Spanish and Cata¬
lan, there is, as has often been pointed out, more common ground
between Spanish and Portuguese than between Spanish and Catalan.
The Catalans have always stressed this fact, because during their long
history of suppression Catalan was not granted the status of a language
in its own right. In the Spanish literature on the subject, in documents,
legal papers, political speeches, and in newspapers, Catalan was often
referred to as a dialect, a vernacular, a variety, (‘una modalidad’) - all
terms that usually implied negative attitudes towards Catalan and re¬
flected a feeling of political superiority on the part of Castilians and
others. Only in times of independence was Catalan called ia lengua
catalana’. On the other hand, terms such as ‘la lengua espanola’, ‘el
idioma espanoT or even ‘cristiano’ (as in ‘hablar en cristiano’ = to
speak Spanish and not some foreign lingo) were applied to Castilian
Spanish only. Today, Catalan is officially one of the languages of
Spain, the others being Basque, Galician and Castilian Spanish; the
latter is one of several varieties of Spanish, the one chosen as the
standard, and also the national language.
13.3.1 A sociolinguistic historical overview

The first documentary evidence of a language recognizable as Catalan
can be traced back to the tenth century, which was the time when the
Catalans gained independence from the Franks and then gradually
266 CASE STUDY II; THE CATALANS
began to win territory from the Moors in the south. Between the
twelfth and fifteenth centuries Catalonia emerged as a politically and
economically influential power. Its territory extended over the whole of
what is today Catalonia (on both sides of the Pyrenees, i.e. including
what is sometimes called Catalunya Nord on the French side), Valen¬
cia, the Balearic Islands and other smaller adjoining areas. By the
twelfth century it had formed a federation with Aragon and eventually
became the Kingdom of Aragon. Catalan gradually replaced Latin as
the official language, and it also became one of the great literary lan¬
guages of the period, producing some of Catalonia’s most famous
writers and poets, thus providing later generations of Catalans with a
‘great tradition’ (Fishman 1971) to refer back to. Such literary devel¬
opment was facilitated by a fairly liberal form of government. The
Catalans and Aragonese had united under the Crown of Aragon on
equal terms, i.e. respecting each other’s legal, cultural and linguistic
habits. However, with the marriage of Isabel de Castilla to Fernando
de Aragon (the successor to the Catalan throne) Catalonia, at the end
of the fifteenth century, became part of the Spanish kingdom. Isabel
claimed for Castile the exclusive right to conquer and colonize Ameri¬
ca, whereas Fernando was committed to expanding Catalonia’s
maritime power in the Mediterranean. The new transatlantic routes
brought immense wealth and power to successive Spanish monarchs
and meant the gradual economic and, as a consequence, political and
cultural decline of Catalonia.
^he 2(X) years of the Habsburg dynasty in Spain (the sixteenth and
seventeenth centuries) presided over the castilianization of the Catalan
nobility and the intellectuals, but it was the eighteenth century that
brought the most repressive measures against Catalan. During the War
of Spanish Succession (1702-1713), which followed the death of the
last Habsburg monarch, Catalonia sided with the Austrians against the
other contender to the Spanish throne, the House of Bourbon. This was
to have disastrous consequences for Catalonia when the Bourbon
Philip V became the next king of Spain and established absolutist and
centralist mle. Catalonia suffered political and cultural repression of a
severity hitherto unknown. The infamous ‘Decretos de Nueva Planta’
(new arrangements) of 1716 abolished all traces of Catalan self-gov¬
ernment and the official status of the Catalan language. Francesc Ferrer
i Girones, previously one of Catalonia’s senators in the Spanish Cortes,
devotes a large part of his book (Ferrer i Girones 1985) on the political
persecution of the Catalan language to the discussion and analysis of
THE CATALAN LANGUAGE
267
eighteenth century documents which either explicitly or by implication

banned the use of Catalan, discriminated against its users or under¬
mined its status as a language. In numerous laws, decrees and
regulations, measures were laid down imposing Castilian as the only offi¬
cial language to be used, both in its spoken and written forms, by those
involved in administration, education, the armed forces and the Church.
Cultural, linguistic and political suppression was to last for more
than 200 years, during which Castilian Spanish was the only language
permitted for public use, in both speaking and writing. The nobility,
higher clergy, military officers and higher civil servants became com¬
pletely castilianized, and monolingual in Castilian Spanish; the middle
classes retained spoken Catalan for informal contact with friends and
relatives, and only the illiterate rural population, with little need or
opportunity for contact with the state authorities, remained monolin¬
gual in Catalan. The language therefore became mainly a spoken
variety, and for many it became associated with rural backwardness.
For a long time a diglossic situation existed, although the domains of
the use of the low variety (Catalan) gradually became extended and
those of the high variety (Castilian Spanish) also shifted. One could
have expected - and this was certainly the intention of eighteenth- and
nineteenth-century politicians in Madrid - that with the rise of the
middle classes, urbanization, improved communication and the spread
of universal education, language use would have moved inexorably to¬
wards Castilian. In the history of other minorities we often see that the
language of the dominant majority sooner or later becomes adopted by
the middle class, thereby ensuring a wide acceptance of its enhanced
status and eventually leading to a major language shift in the area. This
happened, for instance, with the spread of English in Wales and Scot¬
land, of French in Alsace and Brittany, and also in Bmssels, although
the pattern there was slightly different (see Baetens Beardsmore 1983).
However, no such shift occurred in Catalonia.
The nineteenth century saw two important advances in Catalonia:
industrial development and the emergence of Catalanism. The indus¬
trial revolution fed on fertile ground in the region and brought with it a
modem economic infrastructure, increased commercial activity, an ex¬
pansion of the larger towns and considerable prosperity to the area. A
strong urban middle class emerged which was markedly different from
the Spanish-speaking nobility and upper bourgeosie, who were essen¬
tially agriculture-based. This new middle class was liberal in outlook,
enterprising in commerce and industry, and eager to invest and partake
in cultural activities. This, and the influence of the Romantic move¬

ment, fuelled the ‘Renaixenga’ or Catalan cultural renaissance. It
originated in literary and cultural attempts to revive pride in Catalan
achievements during the Middle Ages and, more particularly, in the
language itself. As time went by it developed into a fully fledged cultu¬
ral movement and ultimately fed into Catalanism as a political creed.
The foundation of the Opera House in Barcelona and the famous archi¬
tectural works of, among others, Antoni Gaudi were its physical
manifestations, and literary activities which blossomed particularly in
the second half of the nineteenth century found their expression in the
use of the Catalan language. The use of Catalan as a written medium
was extended tentatively to the field of journalism, and also to private
education. The beginning of the twentieth century was marked by two
milestones for the codification and standardization of the language: the
approval in 1913 of the spelling mles worked out by the Institut d’Es-
tudis Catalans (the Catalan Academy, founded in 1907 and considered
to be the highest authority on matters of language); and the publication
in 1918 of Pompeu Fabra’s Gramatica Normative de la Llengua Cata-
lana (Normative Grammar of the Catalan Language), which was to
become the standard grammar of Catalan.
Towards the end of the nineteenth century, Catalonia’s bourgeoisie
began to embrace Catalanism, i.e. modem Catalan nationalism, as a
political stance as well as a cultural one. The movement demanding
political independence for Catalonia from the Madrid government ga¬
thered momentum. This body of public opinion was now joined by the
upper middle classes, although more cautiously, but as Salvador Giner
points out (Giner 1984), there was never any attempt to reach the
lower echelons of society. The rural areas were more traditional in out¬
look, and largely Catalan-speaking anyway. The working classes, on
the other hand, which were now growing in numbers, owing to urban¬
ization and immigration, became the Castilian-speaking section of
Catalan society and they were more open to non-Catalan political in¬
fluences. Some measure of autonomy was achieved by the creation of
the Mancomunitat in 1914, a political body consisting of repre¬
sentatives from all Catalan regions. The Mancomunitat was especially
active in promoting cultural activities and in education, encouraging
the training of teachers in Catalan and developing new teaching meth¬
ods and educational ideas, such as those inspired by Pestalozzi and
Montessori. In 1915 there were fifteen Montessori and several Pesta¬
lozzi schools in the region, an impressive demonstration of the
progressive educational thinking of Catalan parents at the time.

The use of Catalan in the public sphere received a serious blow
under the dictatorship of General Primo de Rivera (1923-30), who
banned the use of the language in schools and public administration.
But when the Second Republic was proclaimed in 1931 Catalonia was
granted autonomous status. Its moderately left-wing Catalan nationalist
government, the Generalitat (the name was taken from the liberal and
independent government Catalonia had enjoyed in medieval times)
continued to promote the active use of Catalan in all spheres of life,
again making cultural activities and education their priorities. Catalan
was used in administration, the media, commerce and, to an increasing
degree, in schools, universities and training colleges. Private and semi¬
private schools often spearheaded the use of Catalan, as well as the
introduction of educationally progressive methods. Many Castilian¬
speaking immigrants proved willing to learn Catalan, seeing it as the
language of upward social mobility. However, the advance of catalan-
ization, and of Catalanism, often encountered opposition, particularly
from outside Catalonia, and it was slowed down considerably by
clashes between left- and right-wing political forces during the last two
turbulent years of the Second Republic, before the outbreak of the
Spanish Civil War (1936-39).
The fortunes of Catalan deteriorated dramatically after the Civil
War. Again the language was forbidden in public life, as was any
manifestation of Catalanism. The restrictive measures taken by Franco
against Catalonia (dubbed one of the ‘traitor provinces’, the other
being the Basque Country, for their support of the Republican side
during the war) have been described by many sociolinguists, both
Catalan and foreign. All express equal dismay at the vindictiveness of
these policies (e.g. Stephens 1976; Stmbell i Trueta 1985a and b; De
Cicco and Maring 1983; Woolard 1989). The leading intellectual elite
had been forced into exile abroad, and those who were left ‘were im¬
bued with the centralist way of thinking’ (Stmbell i Trueta 1985a).
With regard to language, the axiomatic claim was made that Castilian
Spanish was a world language, the language of true Spaniards, of
economic and political power and prestige, even an ‘imperial’ lan¬
guage. Catalan was branded as a vernacular, a dialect, the mark of the
backward mral peasant. To use it among educated people was tanta¬
mount to a declaration of being anti-government, even anti-Spanish.
Every outward manifestation of Catalan was eliminated in the early
years of the Franco era: the names of towns and villages were castil-
ianized, streets and buildings renamed, publishing houses, bookshops,

public and private libraries searched for Catalan books and those found
destroyed. One of the greatest losses was the priceless collection of
Pompeu Fabra. It was burned in the street. Francesc Vallverdu, one of
Catalonia’s most prominent sociolinguists, refers (Vallverdu 1973) to
the years 1939-50 as the period of ‘persecuted bilingualism’, which
was followed, he says, by a phase of ‘tolerated bilingualism’. Per¬
mission for the publication of certain books and the staging of some
plays in Catalan was granted in the late 1940s. But since censorship
laws applied all over Spain and a large number of titles remained on
the list of books forbidden by Franco’s regime, the range of works
published and plays performed in Catalan was not very wide. Nor were
they the kind that would appeal to those who used Catalan in private,
as most Catalans were opposed to the dictatorship. The ban on the use
of Catalan in the media and in schools remained in force almost to the
end of the Franco period. In general, the last few years of the dictator¬
ship saw a good deal of tolerance in practice, although officially the
restrictions still applied. In the words of John Hooper, for instance, the
Institut d’Estudis Catalans led
a curious, half-tolerated, half-clandestine existence under Franco. It held

weekly meetings, held courses on the language, literature and history of
Catalan in private houses, gave receptions and went so far as to publish
books and pamphlets, some of which were even bought by the government
xfor displays at international exhibitions.
(Hooper 1986: 235).
The Catalans had to wait until 1978 before their language was
granted official status by the new democratic Constitution of Spain, a
number of royal decrees and the 1979 Statute of Autonomy. The Law
of Linguistic Normalization followed, in Catalonia, in 1983 (see later).
However, the restrictive measures against Catalan had lasted long
enough to have a severe impact on the use of and attitudes towards
Catalan. Castilian Spanish had become institutionalized, and many
generations of Catalan schoolchildren had grown up with little oppor¬
tunity of becoming familiar with the language of their parents. More
important still was the impact of those non-Catalans who had immi¬
grated to Catalonia and had neither opportunity nor reason to learn or
use Catalan.
13.3.2 The impact of immigration

Immigrants from Murcia, Aragon and Galicia who came to Catalonia
in the eeu'ly part of the century were able to assimilate over the years
into Catalan society, and in many cases their children learnt Catalan at
school. But the new wave of immigration proved impossible to inte¬
grate, because of the sheer numbers involved and the specific cultural
and social conditions under which it took place. Between 1950 and
1975 approximately two million immigrants settled in Catalonia (many
of them from the south, Andalusia, and again from Galicia), which
meant an increase of 37.15 per cent in the population of the area, as
compared to an average rise of 12.47 per cent in the rest of Spain.
Around 1975 nearly one half (49 per cent) of the people living in the
Municipality of Barcelona were of non-Catalan stock; the figure for the
province of Barcelona excluding the capital was 46 per cent (figures
from Hooper 1986). Catalan towns and cities had considerable diffi¬
culty in coping with the sudden influx of immigrants, and they were
housed in new and poorly built housing estates on the outskirts; just
two of them, Santa Coloma and I’Hospitalet, on the outskirts of Bar¬
celona, now have more than 200,000 inhabitants each, the vast
majority non-Catalan in origin (Strubell i Trueta 1984b). The birth rate
among immigrant families was high (by comparison with the more af¬
fluent urban Catalan society), thus aggravating their poor living
conditions while also adding to the number of non-Catalan speakers.
Socially and educationally, they were more often than not disadvant¬
aged; many had not completed their primary schooling at the time of
immigration, and illiteracy was widespread (Strubell i Trueta 1985b).
The provision of schools, hospitals and social services was totally in¬
adequate in the dormitory towns. The resulting social and linguistic
divisions are marked today by resistance to linguistic and cultural inte¬
gration, by juvenile delinquency and by other signs of social
disintegration which are aggravated further by the level of unemploy¬
ment (the highest in western Europe) that exists everywhere in Spain.
Contact with the Catalan-speaking population has been, and still is,
very limited. It has been described as contact with people one or more
rungs higher up on the social ladder, like the doctor, priest, teacher and
civil servant. Some have interpreted it as a positive type of contact,
since it may foster the desire on the part of immigrants to learn Catalan
so as to participate more fully in society, or at least to have their child¬
ren taught Catalan at school. Thus, the reason most often given by
those asked why they wanted their children to learn Catalan was ‘that
it is necessary in order to get on in life’. Such expectations, however,

indicate that parental motivation towards linguistic integration is purely
instrumental (i.e. utilitarian) and extends only as far as their children.
For language learning and assimilation into Catalan society to be suc¬
cessful, immigrants, and particularly second-generation immigrants,
need to be positively motivated themselves, driven by the attractions
that they themselves see in becoming fully Catalan. But just what these
attractions are is not always easy to detect.
Social inequality, marginalization and, above all, a rapid increase in
numbers have contributed to the fact that Catalonia’s immigrant popu¬
lation has acquired political and social significance. They are today the
strongest force that poses a threat to the Generalitafs attempts at cata-
lanization, since the Catalan government depends on the immigrant
vote. The immigrants are well aware of their political importance, and
they know that they can resist catalanization if it becomes too unpalat¬
able. Just which path they will take, whether they will cooperate fully
or try to develop their own social and cultural ambitions separately, is
not yet clear. Much depends on the Generalitafs power and skill in
controlling the media and the education system, on the one hand, and
in finding solutions to the real social problems and divisions on the
other, while still pursuing at the same time truly democratic, tolerant
and pluralist policies. It is, obviously, not an easy task.
13.3.3 The present linguistic situation

X
The two most influential factors bearing directly on a minority lan¬
guage are its political status and the socio-economic position of its
speakers. As regards the latter, Catalan has traditionally enjoyed the
full support of both the middle classes and the rural population; in
addition, by the beginning of this century large sections of the upper
class, and also of the working classes, spoke Catalan. Naturally, fewer
people used Catalan during the forty years of the Franco dictatorship,
but at the same time Catalan became the language associated with pol¬
itical resistance, and began for this reason to appeal to many people
from all sections of the population. As the speakers of Catalan are
nowadays regarded positively, and their socio-economic status and
educational background are generally higher than those of Spanish¬
speaking immigrants, their language, Catalan, also enjoys considerable
prestige. This has caused the language to be considered to be appropri¬
ate for communication in any situation and in connection with any
topic.* Today the language choice of a Catalan-Castilian bilingual is
LANGUAGE PLANNING SINCE 1975 273
based rather more on the language of her or his interlocutor, and not so
much on the formality of the situation or on the medium (i.e. written or
spoken language) employed. Code-switching in Catalonia does not
necessarily occur in one direction, from Catalan to Castilian; it may be
in either direction. The diglossic situation that was, in the past, at¬
tributed to Catalonia, has ceased to exist.
Azevedo (1984) observed that bilingual conversations seemed to
have become the rule rather than the exception in Catalonia - just as
passive bilingualism has, too. However, the present linguistic situation
in Catalonia cannot be adequately described with a single label. Sev¬
eral forms of bilingualism coexist among the population of the area,
ranging from active to passive, and from incipient to balanced. But
officially the term ‘bilingualism’ does not occur anywhere. The lan¬
guage policy of the Generalitat is geared towards ‘restoring’ Catalan,
not towards establishing bilingualism. Bilingualism is a de facto phe¬
nomenon among Catalan speakers, but it is not usually openly
advocated. Baetens Beardsmore’s (1983) term of ‘bilinguisme de resig¬
nation’ may not be entirely inappropriate here, although perhaps many
Catalans would look at it as ‘involuntary bilingualism’, as they prefer
to use their own language only; but if required to speak Spanish it
becomes clear that they are fully bilingual.
13.4 Language planning since 1975
The legal framework for the Generalitat's language policies is pro¬

vided by three major pieces of legislation: the Constitution of 1978, the
Statute of Autonomy of 1979 and, most extensively, the 1983 Law of
Linguistic Normalization in Catalonia. In the last years of the Franco
era and during the period of the transition to democracy, decrees and
regulations were passed which began to show some embryonic recog¬
nition of Spain’s ‘regional features’, including its languages (terms
such as ‘peculiaridades regionales’, ‘lengua nativa’ and ‘lenguas nacio-
nales’ were used); and it was recommended that these languages
should, at least on a voluntary basis, be included in school curricula.
Little official action was taken in some non-Castilian-speaking areas,
however, because of a lack of local organization and resources. In
Catalonia preparations were under way, though. The reinstatement of
the Generalitat in September, 1977 (it remained provisional until the
first general election under the new constitution was held in 1979) was
an important factor. No time \Yas lost in forming from among its ex¬
ecutive council a Council for Education and Culture which in turn set
up the Catalan Language Service (‘Servei del Catala’), charged with
the evaluation and coordination of all aspects of teaching of Catalan
and through Catalan. In cooperation with the Ministry of Education in
Madrid, the reorganization of the curriculum and the teaching profes¬
sion was also begun, so as to include the teaching of Catalan and the
establishment of the necessary new posts.
The reason why the Generalitat made its language policy a priority
was the realization that Catalan was facing a crisis. The Franco legacy,
plus the influx of immigrants, presented, as we have seen, a dual threat
to the language. Before 1975 no official figures existed about the num¬
ber of Catalan speakers, since the censuses did not include questions
on language. According to a report published by the Spanish Govern¬
ment in 1975, 71 per cent of the population in the principality could
speak Catalan, although the proportion of those who actulaly did speak
it was somewhat lower. A sharp division between the capital and the
provinces of Lleida, Tarragona and Girona became evident: in Barcelo¬
na only 39 per cent of the inhabitants were Catalan speakers, whereas
the figure for the provinces was 90 per cent. Subsequent research
showed that many speakers of Catalan could neither read nor write the
language. A further problem was highlighted when a report was pub¬
lished showing that Catalan tended to be used more often as the
language of social interaction, i.e. in conversation and at work, than as
the language of the home. In mixed marriages the use of Castilian
Spanish was often preferred to that of Catalan - and, as is well known
(and as the example of Alsace showed), once a minority language
ceases to be the language of the family, its continued existence is seri¬
ously threatened. Other signs indicating a weakening of Catalan
vis-d-vis Castilian Spanish were seen in the increasing number of loans
and castilianisms apparent even in the language of educated speakers
and writers, a problem largely attributed to the prolonged linguistic
accommodation Catalans had been forced to undergo.
Figures published in 1980 showing the knowledge of Catalan among
the inhabitants of the region were not very encouraging. They referred to
the population as a whole and, taken together with the figures for the
school population’s home language, they confirmed the impression that
the main thmst of local government language policy had to focus on the
catalanization of the education system, if any real change in language use
was to be achieved:
LANGUAGE PLANNING SINCE 1975 275
TABLE 13.1 Percentage of population by knowledge of Catalan;
Percentage
No data 2%
Castilian speakers who do not understand Catalan 18%
Castilian speakers who understand Catalan 12%
Castilian speakers who speak Catalan 10%
Perfect bilinguals 6%
Native speakers of Catalan 52%
Total 100%
Source: Subirats 1980 quoted in Sabater 1984
Unfortunately, no further elucidation is offered as to the precise

meaning of the category ‘perfect bilinguals’. In more recent publica¬
tions the term has been dropped altogether and language knowledge is
indicated in terms of reading, writing, speaking and understanding.
Table 13.2 gives some information on language use.
TABLE 13.2 Language of pupils in Catalonia (percentage):
In general basic In higher secondary

education (compulsory education (15-18 years)
schooling 6-14 years)
Baccalaureate Vocational
training
Catalan in 33.71 47.86 33.73

the family
Bilingual 12.64 12.83 12.06

families
Castilian 53.65 39.31 54.21

Spanish in
the family
Source: Generalitat de Catalunya Departament d’Ensenyament 1982

The figures for the groups of youngsters who speak Catalan are
particularly interesting, as they clearly show that this language is
spoken most often in families who value education (as they keep their
children at school beyond the age of fifteen). But they also reveal that
the use of Catalan within the younger group, i.e. those who were bom
around the end of the Franco era, is quite low.
Article 3 of the Statute of Autonomy constitutes the basis for the
Generalitat’s language policy. It states that Catalan is ‘Catalonia’s own
language’, it spells out its co-official status and it asserts the Generali-
tat's determination to ensure the normal and official use of both
Catalan and Castilian Spanish. It also makes the Generalitat respon¬
sible for language policies. The means by which Catalonia’s
government intended to carry out these policies were laid down in the
Law of Linguistic Normalization. The term ‘normalization’ reflects the
underlying philosophy of Catalan language planning, which is ex¬
pressed thus in the preamble of that Law:
The restoration of Catalan to its rightful place as Catalonia’s own language

is the unquestionable right and duty of the Catalan people and must be pro¬
tected and respected. In this regard, knowledge of the language must
spread throughout the whole of Catalan society, to all citizens regardless of
the language they normally speak, within a global framework in which
everyone will accept the use of both languages and recognize and contrib¬
ute to the recovery of Catalan as one of the fundamental aspects of the
reconstruction of Catalonia.
The main objective was thus to overcome what was referred to as ‘a

situation of linguistic inequality’. The radical nature of this aim con¬
sists of the idea that the whole of the population of Catalonia should
learn Catalan, not only those who are self-professed Catalans. It is of
course this aspect of Catalonia’s language policy that is the most con¬
troversial and even, possibly, illusory. For, whereas it is relatively easy
to place Catalans in a position to use their language in all walks of life
and to have their children educated in the language of their forefathers,
it is more difficult to succeed in turning almost half the population,
which is mainly of non-Catalan descent, into speakers of Catalan. On
the other hand, it may well be the case that only aggressively radical
language policies stand any chance of arresting, or perhaps reversing,
the decline of a language.
In 1980 a Directorate General of Language Policy was set up and
given the task of spreading the use and knowledge of Catalan; it was
THE PUBLIC USE OF CATALAN 277
also made ultimately responsible for language planning. Today a num¬

ber of bodies are involved in the various activities which are being
carried out under the Law of Linguistic Normalization, such as provid¬
ing special-purpose Catalan classes for state employees, developing
technical vocabulary for government departments and the system of
administration of justice, and promoting Catalan in publishing, the
media and, most importantly, in education.
The ‘Campaign for Language Normalization’ (under the slogan ‘el
catala, cosa de tots’ - freely translated, ‘[learning] Catalan concerns us
all’) is an ongoing concern in Catalonia. It takes the shape of, for
example, slogans and short features published in the newspapers, and
posters and banners displayed in public places, and radio and television
broadcasts. The aim is to increase language consciousness with a view
to encouraging Catalans, and all the inhabitants of Catalonia, to use
Catalan more frequently and consistently, and to foster positive atti¬
tudes towards this language. The campaign extends to other activities
that relate to the spreading of the use of Catalan. As stipulated in the
Normalization Law, certain bodies may become eligible for tax deduc¬
tions, or even total exemption, by switching to the use of Catalan (e.g.
within a firm), which obviously can make support for the Generalitat’s
language policies an attractive proposition even for the less ideologi¬
cally inclined world of business and commerce.
Naturally, the efforts to re-establish Catalan in all domains of public
life have met with some scepticism in certain quarters and, at times,
have caused politically or ideologically inspired protests by non-Cata¬
lans both inside Catalonia and in other parts of Spain. But there is no
evidence that the rights of Castilian speakers have not been respected.
Azevedo (1984) points out that many Catalan linguists and educationa¬
lists were among the first to show concern for those rights; and he
adds: ‘Actually, official efforts to re-establish the use of Catalan have
been rather tactful, judging from the general tone of the 1982 Linguis¬
tic Normalization Campaign’ (Azevedo 1984: 323).
13.5 The public use of Catalan
The Law of Linguistic Normalization was passed in 1983 in order to

promote the ‘social presence’ of Catalan. It contains provisions for the
use of Catalan in administration, the media and education, and it guar¬
antees institutional support by the Generalitat. A general overview of
these provisions, and of the impact they have had on the use of Catalan
in public life, is given in the following sections.
(1) Administration Catalan is the language of the Generalitat and of

the Catalan territorial administration, of local government and of those
bodies that come under the Generalitat. New state employees are re¬
quired to demonstrate a knowledge of Catalan, and those who were in
employment before the linguistic changes took place are encouraged to
attend language classes in Catalan. These are state-subsidized and often
held during working hours, thus complying with the pledge that lin¬
guistic normalization would be carried out without discrimination
against non-Catalan speakers. While internal communication in gov¬
ernment departments and offices tends to be in Catalan, Castilian
Spanish can be used on request. The public have the right to conduct
their dealings with the authorities in the language of their choice, and
all official material is published in both languages. It is therefore not
too optimistic to assume that, within a relatively short period of time,
most of the administration business will have been catalanized. It is
more difficult to predict a similar situation in the media and in educa¬
tion, which are less easily controlled by policies, as they are subject to
the influence of market forces, social factors and psychological vari¬
ables - and therefore provide an enormous challenge to language
planners.
(2) The media The Generalitat's role in catalanizing the mass media is
lar^ly restricted to subsidizing the publication of newspapers, peri¬
odicals and magazines which are partially or entirely written in
Catalan, and to providing some financial aid for the production of
plays and the dubbing of non-Catalan films. There is only one daily
newspaper {Avui) published entirely in Catalan, however, and in 1986,
ten years after it was first launched, it had a circulation of 40,232
copies. There are also a number of local papers, ranging from those
printed wholly in Catalan to some which contain only a small Catalan
contribution.
There are now several officially-run radio stations covering Catalo¬
nia that are subsidized by the Generalitat and broadcast only in
Catalan. Between them they cover all kinds of programmes, from news
and sports to current affairs and music. There are, in addition, numer¬
ous private local radio stations operating partially or wholly in Catalan,
frequently supported by town councils. The 17 biggest stations offered
24 per cent of air-time in Catalan, 9 per cent in both Catalan and Cas-
tillian and 67 per cent in Castillian (figures for 1986).
In 1984 a Catalan television channel, TVS, was started, and by the
following year it was broadcasting some forty-four hours per week in
Catalan, and in 1988 the figure had reached 300, with peak audience
figures being achieved in the early afternoon (i.e. Spanish lunch-time,
when normally all members of the family have a meal at home) and
late evening. It is popular also because of its extensive news, its serials
(or ‘soaps’)^ and (especially) sports coverage. It receives its income in
part from the Generalitat and partly from TV advertising.
The increased popularity of Catalan radio and television is clearly
an encouraging sign to all those involved in the recovery of Catalan, as
the use of the language in the media increases awareness and knowl¬
edge of Catalan. The problem that arises here, as in administration, is
that of finding enough people who speak grammatically and stylisti¬
cally correct Catalan, so that they can provide a good linguistic model.
Journalists and broadcasters are not necessarily as competent linguisti¬
cally as professional linguists would like them to be, and their speech
has been criticized for being phonetically and lexically castilianized,
which highlights yet another area in which standardization of usage
still has to take place.
Publishing in Catalonia has a long and distinguished tradition, and
there are more than a hundred publishing houses in Barcelona, most of
them publishing in both Catalan and Castilian Spanish, often with one
part of the enterprise (the Castilian) supporting the other. During the
Franco years publication in Catalan was seriously restricted, but never¬
theless kept alive by dedicated authors and publishers who often
produced books at their own expense. When, in 1977, the then prime
minister, Adolfo Suarez (a native of Castile), asked somewhat scathing¬
ly in a foreign publication, ‘How could we write a book on nuclear
physics in Catalan?’ (Fabre 1979), the response was immediate: a
whole series of books was published, on nuclear physics and a range of
other highly technical subjects. Book publishing nowadays receives
financial help from local funds, through the ‘Serve! del Llibre’ (Book
Office), which subsidizes the publication of some 200-300 works of
fiction and non-fiction each year. Copies of these books are then made
available to the Generalitat and its public libraries. The number of tit¬
les in Catalan is said (in an official publication by the Generalitat,
‘Social Communication in Catalonia’, 1988) to have increased from
2175 in 1982 to 4145 in 1987. The number of copies sold overall grew
280 CASE STUDY 11: THE CATALANS
from 3.3 million in 1980 to 5.3 million in 1984, even though the aver¬
age sale per book dropped by 15 per cent. With all three levels of
education heavily involved in the catalanization programme, there is
still a growing demand from the educational system for books in Cata¬
lan. Nevertheless, publishing is an uncertain and expensive business
everywhere. In Catalonia it remains heavily dependent on sustained
government aid and private support.
(3) Education It has long been recognized that unless Catalan is taught
in schools and eventually becomes the medium of instruction, all the
other efforts towards language recovery will remain cosmetic and may
be short-lived. The adoption of Catalan as a medium of instruction is a
challenging task, and it offers the opportunity of undertaking both lin¬
guistic and curriculum reform. Many committed Catalan politicians
and educationalists have approached the enterprise with a great deal of
enthusiasm. For them, the aim of language ‘normalization’, ‘recovery’
and restoration in the school context is: ‘to put Catalan pupils who do
not know their own language (but are familiar with another one)
through an immersion programme within a linguistic framework which
cannot be called that of a minority.’ (Abeya 1985).
But at the same time language policy in education is a highly sensi¬
tive issue. The Catalan Government is faced with the task of
overcoming the considerable problems posed by little interest and even
hostility on the part of parents and teachers - not to mention the atti¬
tude A)f the remaining Spanish population, the reservations of central
government, and a general lack of resources, both human and material.
Section II of the Law of Linguistic Normalization (1983) includes
seven Articles which deal with the question of language policy in edu¬
cation. The first one simply (and rather ambitiously) states: ‘Catalan,
as Catalonia s own language, is also the language of education at all
levels’. Catalan is obligatory in schools, and all children need to attain
an acceptable level of proficiency by the end of their basic education
(i.e. by the time they are fourteen years old). Only if children can
demonstrate adequate mastery of Catalan will they be awarded the
Certificate of General Basic Education (school-leaving certificate). The
principle that children are entitled to receive their early instruction in
the mother tongue is, however, accepted, so that upon entering school
children may be taught initially either in Catalan or in Castilian Span¬
ish. The Act lays down that children should not be segregated into
different schools purely on the basis of their first language, and it adds
that Catalan should be used progressively as the medium of instruction

as pupils’ mastery of it increases.
In higher education, staff and students are free to choose which lan¬
guage they wish to use, but all centres at tertiary level are obliged to
offer language courses for non-Catalan speakers. As far as teacher¬
training is concerned, the Act is quite specific: ‘All teachers must be
proficient in both official languages’. This requirement means that the
curriculum for teacher-training courses has to be designed in such a
way as to ensure that students are trained to teach their subjects in both
languages.
It will be some time before the provisions of the law are fully im¬
plemented, particularly in the public-sector secondary schools. In the
private sector, catalanization has progressed considerably faster, but
then private Catalan schools have a long tradition of promoting Catalan
culture, including the language. Some 40 per cent of all schools in
Catalonia are private. They have to conform to the general educational
objectives outlined by the Ministry of Education in Madrid, but they
are otherwise free to follow particular religious or pedagogical princi¬
ples. Although school fees are not generally as high as they can be in
other countries, private schools in Spain are normally attended only by
middle- and upper-class children, which means that, once again, it is
the upper layers of society that are spearheading catalanization - but
also contributing to linguistic, as well as social, segregation (see sec¬
tion 13.3.3).
The linguistic normalization programme in schools is faced with
two major problems. One consists in the uneven distribution of the
Catalan- and Castilian Spanish-speaking population. In some parts of
the Barcelona metropolitan area (i.e. the districts with the heaviest con¬
centration of immigrants), more than 80 per cent of inhabitants only
speak Spanish; in such areas the need to learn Catalan is perceived as
being at best marginal. These are also socially deprived areas, with
high levels of unemployment. Only an improvement in the quality of
public services, coupled with a higher degree of social integration be¬
tween Catalan and Castilian Spanish speakers will, one feels, ensure
the successful catalanization of schools in these areas. The other prob¬
lem is caused by a scarcity of qualified staff able to use Catalan as the
medium of instruction: only about half the teachers who work in Cata¬
lonia are of Catalan origin, and the number of those who have
undergone teacher training in Catalan is smaller still. A further factor
militating against linguistic reorganization in schools is the poor
knowledge of Catalan, even among teachers (Amau and Boada 1986).

Since 1978 the Generalitat has assumed most responsibilities for
education in Catalonia, including the appointment of teaching staff
(previously controlled centrally from Madrid) and retraining in Cata¬
lan. Another important measure was the setting up of the Education
Department’s Office for the Teaching of Catalan (Serve! d’Ensen-
yament del Catala SEDEC), entmsted with promoting and organizing
the teaching of, and in, Catalan in schools. This body supplies special¬
ist Catalan teachers to schools that request them, it monitors the use of
Catalan in schools and holds seminars for parents’ and teachers’ associ¬
ations. Parental support and involvement is clearly a vital aspect if
linguistic recovery is to succeed. SEDEC and a number of other agen¬
cies regularly carry out surveys to collect information on the spread
and use of Catalan, both in schools and in other fields. The results of a
survey among the school population carried out in 1981-82 show that
the implementation of any aspect of linguistic legislation has to be dis¬
cretional in some areas of Catalonia, as it must take into account such
factors as the linguistic background of pupils, the teachers’ knowledge
of Catalan and variations in the local quality of the teaching services.
The survey covered a total of 3399 schools, of which 1777 were pri¬
vate- and 1622 public-sector ones. In all schools Catalan was taught as
a subject, and in 65 per cent of them instruction was either wholly or
partially in Catalan. This figure may seem high, but it must be seen in
the light of what was said earlier about the prominence of Catalan in
private schools.
With reference to the linguistic background of pupils, the SEDEC
report showed that
59.52 per cent of pupils could speak Catalan;

24.88 per cent of pupils understood Catalan;
15.60 per cent of pupils did not understand Catalan.
The 60 per cent or so of children who spoke Catalan were further

classified into three groups:
(1) monolingual Catalan speakers 33.7 per cent;

(2) bilingual Catalan/Spanish speakers 12.6 per cent;
(3) speakers of Catalan as second language 53.7 per cent.
With regard to teachers knowledge of Catalan, the survey indicated

that the proportion of those speaking Catalan was 67.73 per cent -
slightly higher than that of the pupils. Of the 32.27 per cent of teachers
who said that they did not speak Catalan, 76.94 per cent claimed to
understand it, which shows that the proportion of teachers who could
not understand Catalan was actually higher than the percentage of pu¬
pils in the same position (23.06 per cent as against 15.60 per cent).
This discrepancy may, however, be attributed to children’s and adults’
different views of what ‘understanding’ means.
The geographical distribution of Catalan- and Castilian Spanish¬
speaking pupils is, not surprisingly, similar to that of the population as
a whole. There is considerable variation between rural and urban dis¬
tricts, and between those with higher or lower concentrations of
immigrants. Figures from Siguan (1980) show that, for the whole of
Catalonia, some 50 to 55 per cent of pupils had Catalan as their home
language. The geographical variation ranged from almost 100 per cent
in rural areas to 47 per cent in Barcelona and just 27 per cent in the
industrial outskirts of the capital. It is, incidentally, in these areas that
some of the major school projects (immersion courses specially de¬
signed and carefully monitored) are to be found.
(4) School programmes Since the incorporation (1978) of Catalan in

schools, either as a subject or as the language of instruction, three dif¬
ferent models have been used:
Model A: total immersion;

Model B: teaching partially in Catalan;
Model C: teaching progressively more in Catalan
(i.e. starting with very little instruction in Catalan and
gradually increasing it until it outweighs Castilian Span¬
ish).
Generally it has been up to individual schools to choose the model

they want to follow, on the basis of the qualifications and commitment
of their staff rather than according to the linguistic background of their
pupils.
Amau and Boada (1986) looked at the development of different
models in schools and at the linguistic abilities and qualifications of
teachers. They found that, of the three models mentioned, A had been
chosen by a small but stable number (8 per cent) of public-sector
schools, which had been following the General Basic Education pro¬
gramme totally in Catalan. In 1982 there was a marked increase, in
relation to previous years, in the number of state schools following
Model B, and the figures for 1983-84 begin to show an increase in the
proportion of schools adopting Model C. Altogether, during the school
year 1987-88 almost 70 per cent of schools in the public sector were
using Catalan wholly, partially or progressively, and in the remaining
schools Catalan was taught as a school subject. Compared to the figure
of 3 per cent for 1975, when Catalan was only ‘tolerated’, these figures
look encouraging, although it is too early to be more than cautiously
optimistic about the spread of Catalan among the school population. In
contrast to the Welsh and Canadian experience of immersion education
the Catalans need to bear one vital difference in mind: their target
groups are not children from motivated middle-class homes. They are,
for the most, the children of immigrants who live in the poorer work-
ing-class districts; and when they leave the school and the immersion
programme there will be little need, or motivation, for them to speak
Catalan.
In their assessment of the factors believed to affect school achieve¬
ment in Catalan and in Castilian Spanish, Amau and Boada (1986)
found that general learning ability was of paramount importance. But
the type of school model adopted was important, too, at least for the
successful establishment of Catalan. Only the total and progressive im¬
mersion models (A and C) provided the necessary linguistic orientation
for children whose family and social environment was predominantly
Castilian.
13.6 The extent of Catalan knowledge
A good deal of statistical data on language issues can now be obtained

from the various local government agencies concerned with the promo¬
tion of Catalan, based on regular surveys in different parts of the
region and on census information. The 1986 census contained several
entries which, taken together, provide an indication of people’s lan¬
guage background and use of Catalan. Also, six specific questions
were asked about the passive and active use of Catalan, and its spoken
and written forms. The following table gives a general overview of the
extent of the knowledge of Catalan in the four provinces of Catalonia
for 1986:
THE EXTENT OF CATALAN KNOWLEDGE 285
TABLE 13.3 Knowledge of Catalan:
Understand Speak Read Write
(%) (%) (%) (%)
Total for
Catalonia 90.3 64 60.5 31.5
Barcelona 89 59.8 58.2 30.1
Girona 95.1 80.1 70.7 39.3
Lleida 96.3 82.8 71.6 37.1
Tarragona 92.8 72.9 63.7 32
Source: Padrons municipals d’habitants de Catalunya, 1986:

Coneixement del catald, CIDC, Barcelona 1987, p.l5
Note - Barcelona is by far the most significant province numerically,

with over 4.5 million inhabitants. The other three together have a
population of about 1.4 million people. These statistics refer to inhabi¬
tants over two years old.
There is a great deal of variation in the distribution of figures if

they are broken down into smaller districts. In the Barcelona areas with
a high proportion of immigrant population (Santa Coloma de
Gramenet, for instance) only a few people are able to read and write
Catalan:
TABLE 13.3a Knowledge of Catalan in Santa Coloma
Understand Speak Read Write

(%) (%) (%) (%)
76 28.6 28.1 11.8
Source: ibid. p. 26
In general terms, during the five years 1981—86, knowledge of Cata¬

lan rose in this same area (Santa Coloma) by 25 per cent. The figures
showing the increase in understanding alone, for the whole of Catalo¬

nia, between 1981 and 1986 are as follows:
TABLE 13.4 Understanding of Catalan
1986 1981 Difference

(%) (%) (%)
Barcelona 89 79.8 +11.9

Girona 95.1 90.7 +4.4
Lleida 96.3 91.8 +4.5
Tarragona 92.8 86 +6.9
Total for
Catalonia: 90.3 79.8 +10.5
Source: ibid. p. 43
An examination of the data available according to age of the popu¬

lation suggests that the Linguistic Normalization Campaign has been
particularly successful in the field of education. Knowledge of Catalan
among the school population and those who have recently left school
has risen quite dramatically: 78 per cent of young people in Catalonia
between ten and nineteen years old speak Catalan (as against 64 per
cent for the population as a whole). The point has been made, however,
that oral ability does not necessarily mean fully effective use of the
language (Stmbell i Trueta 1988).
13.7 The future of Catalan
It is worth reminding oneself of some of the features which distinguish

the Catalan situation from that of other European minorities. The im¬
portance of the long history of involvement of the urban middle classes
in the promotion of Catalan language and culture on a public and pri¬
vate level cannot be underestimated. The bourgeoisie was a driving
force in making widespread use of Catalan socially acceptable, and it
went hand-in-hand with their advancement in the fields of industry,
commerce, education and culture, and later also in political affairs.
Any attempts to stigmatize the language came from outside Catalan
THE FUTURE OF CATALAN 287
society and never really penetrated it ideologically. Castilian Spanish

was imposed in such a crude way, that Catalans never ceased to feel
antagonism towards Castilian administrators, members of the armed
forces and the police. Castilian Spanish was associated with political
oppression and, during the Franco period, with fascism, whereas Cata¬
lan was seen as representing resistance to reactionary centralist forces.
Early standardization of the language and a rich literary heritage also
contributed to the stabilization and prestige of the language. While
Corsicans or Bretons might compensate for their linguistic inferiority
complex by displaying ardent nationalism and demanding political sep¬
aration, the Catalans have concentrated their efforts on pursuing
cultural and linguistic ends. This is not to say that nowadays in Catalo¬
nia there are no demands for political independence from the rest of
Spain. But the proverbial Catalan ‘seny’ (common sense) seems to
have accepted the present quasi-federal solution. There are of course
other likely reasons for this relative contentment. One may be that,
since Catalans cannot reasonably claim to be ethnically very different
from other Spaniards (as can, to some extent, the Basques), political
separation does not appear to many to be a particularly desirable goal.
The financial and ideological commitment of many Catalans to re¬
cover their language is considerable. Indeed, the far-reaching measures
undertaken by the Generalitat's language planners are without equal
anywhere in Europe. In the fields of administration and communica¬
tion, the policy of conscious catalanization has already brought
encouraging results. It is still too early to see what effects endeavours
at school level will have. But it is clear that the burden placed on the
education system in terms of expectations and demands is enormous,
as it is not adequately equipped at the moment, either financially or
pedagogically, to make all pupils competent in Catalan. In any case,
making the whole of Catalonia Catalan-speaking cannot be entirely up
to the schools: a fundamental social change is necessary as well, in the
form of complete social integration and the support of the entire popu¬
lation of the area for the linguistic normalization programme. This can
come about only as a result of changing attitudes towards each other
on the part of both linguistic groups. In this field schools can play a
crucial role (as the Canadian experience of bilingual education has
shown), but other changes have to take place at the same time: one
cannot assume that positive attitudes towards Catalan, or linguistic
skills acquired at school, will last a long time if young people, once
they have left school, no longer feel the need, or have the motivation.
to use the language regularly, and more importantly pass it on to their

children.
The Catalans have been keen to learn from language planning in
other countries, notably Wales and Canada. Despite some similarities,
however, their linguistic situation is not really comparable. In the con¬
text of Spain as a whole, Castilian Spanish-speakers are the majority
and Catalan speakers the minority, but within Catalonia the picture is
reversed, with neither group possessing all the typical traits of either a
majority or a minority. Whether its language policies will make it
possible for the present comunidad autonoma to recover Catalan’s lost
ground, resulting in a change in linguistic behaviour among Catalans
of Castilian descent, remains to be seen. Achieving these objectives
would virtually amount to a sociolinguistic miracle in modem Europe.
Notes
1. There are several problems associated with the standardized ver¬
sion of Catalan: as it is modelled on literary language, it is not
always well equipped to be an effective and modem public me¬
dium. It has been claimed that the norms laid down by rigorous
reformers are at variance with the living language. As a result,
influences from contemporary Spanish are difficult to eradicate
fron everyday usage. Another problem arises from the various
disagreements about the use or non-use of regional variants of
Catalan, particularly in education. For a general discussion and
references, see Azevedo (1984).
2. This low figure is probably more a reflection of the paper’s

quality than the Catalan’s inherent interest in a Catalan news¬
paper. The left-of-centre daily El Pais (53,303 copies, 1986) is
published in Madrid and Barcelona. In the Barcelona edition the
literary supplement and the crossword are in Catalan. Several
times more copies are sold of La Vanguardia, a centre-right daily
(194,553 copies, 1986). In its Catalan edition some material is
always in Catalan.
1. One of TV3’s real successes was that it acquired the screening

rights for the American soap opera series ‘Dallas’ before the
other Spanish networks could do so!
Chapter 14
Case Study III: Migrant workers in the

Federal Republic of Germany*
*The data referred to in this chapter on the Federal Republic of

Germany was prepared prior to German reunification in 1990.
14.1 Old and new minorities
14.1.1 Indigenous minorities

There are today only two long-established ethnic minorities in the
Federal Republic of Germany, and both are to be found in the north of
the country.^
There are some 20,000 members of the Danish minority association
{Den Sydslesvigsk Forening) and an estimated 50,000 speakers of Danish
in Schleswig-Holstein (Walker 1987). The Danish minority enjoys special
legal rights accorded to it by the German state (just as the German mi¬
nority does in Denmark). The minority is well-organized. It sends two
representatives to the state parliament in Kiel (the capital of the Federal
State of Schleswig-Holstein), and it has kindergartens and schools which
receive financial aid from the Schleswig-Holstein authorities, an active
Danish Church and a network of Danish health, hbrary and youth associ¬
ation services. Another reason why their standing is enhanced is their
proximity to the country where their language has official status.
The Frisians, on the other hand, are fewer in number, with some
8000-10,000 speakers (Walker 1987). They have no legal rights, they
are geographically more dispersed, many gave up their language sev¬
eral generations ago, and they have an undeveloped infrastructure.
They live, geographically speaking, on the fringes of Germany, along
the coast of the North Sea and on the Frisian Islands, a considerable
distance away from their brothers and sisters (who could offer them
support) in Friesland, in the Netherlands. The German Frisians do re¬
ceive some financial aid now for the teaching of their language from
the regional government in Kiel, but funds do not extend to the provi-
290 CASE STUDY IH: MIGRANT WORKERS IN THE FEDERAL
REPUBLIC OF GERMANY
>ANE?
•JOR
FRISIANS
"j
^ Hamburg
'•N
I '-N
/
f
\
1 ► Berlin
S
<
r-' . r-'
s
1 Duisburg
)
/> ^Diisseldorf
f
*Remscheid
K Cologne
■>
■s
1
I Frankfurt
\
NX
(
C Mannheim
Ludwigsburg
•Heilbronn
• Stuttgart
Munich
14.1 Map showing the migration and imigration to west Germany

<— denotes countries of origin of migrants
<= denotes countries of origin of German minority members who have
emigrated
sion of teaching materials. Walker (1987) describes the status of Fri¬

sians as ‘symbolically high’ but ‘practically low’, which can be
explained, at least partly, by its strong resemblance to Plattdeutsch
(Ixow German), the varieties of German spoken throughout North
Schleswig and other parts of northern Germany.
OLD AND NEW MINORITIES 291
14.1.2 Refugees and returning German minorities

The small number of indigenous minorities in the Federal Republic is a
consequence of the fact that this state is of recent creation (1949). The
reorganization of Germany into two separate states and the alteration
of virtually all its previous borders meant that those minorities which
had been, at one time, part of the German Empire remained outside the
territory of west Germany. As a result of the Second World War and
political changes in its aftermath, particularly in eastern Europe, large
numbers of refugees came to west Germany (some 12 million before
1961; 1.7 million after 1961, the year the Berlin Wall went up). The
majority of them were German-speakers, but many also spoke a second
language. A good proportion of the latter, particularly those who came
from the Baltic republics, endeavoured to maintain their languages, and
some were able to form well-organized systems of summer schools and
associations. All those who went to west Germany as refugees have by
now become integrated into the host society. The people who kept the
Estonian, Latvian and Lithuanian languages alive in their families are
now too dispersed and small in number even to be considered a mi¬
nority.
A further consequence of the Second World War has led to a more
recent development: the ‘re-emigration’ into the Eederal Republic of
German minority members from eastern European countries, particu¬
larly Poland, Romania and the Soviet Union. Since the 1960s some
2 million people have settled in west Germany and many more con¬
tinue to come. The absorption of this influx of people has not been
easy as it has created a host of economic and social problems. It has
also confronted the German nation with some interesting questions
about what it means to be German. A large proportion of these Aus-
siedler (‘returnees’), as they are called, do not speak German, and
many others have maintained this language only as a spoken medium.
The Federal State offers them language tuition and makes special pro¬
vision for their children, who not only need to learn German for their
schooling, but also find themselves having to adjust to a different edu¬
cation system. Some would argue that they are entitled to the
opportunity to maintain their other language, a view held by the volun¬
tary organizations which are also involved in teaching them German.
But the Federal German state is reluctant to grant this request.
The situation, therefore, is one in which the indigenous population
has had to learn to accept these newcomers as Germans, rather than
‘Poles’, ‘Russians’ or ‘Romanians’, and the Aussiedler, on the other
292 CASE STUDY III: MIGRANT WORKERS IN THE FEDERAL
REPUBLIC OF GERMANY
hand, have been forced to undergo considerable adjustment. Their posi¬

tion has been compared to that of the post-war refugees, but the
difference is that, whereas the refugees had at least lived among Ger¬
mans, these people have not. The German minority in Poland, for
example, has had little or no contact with Germans over the previous
forty or fifty years, even though the German Democratic Republic was
relatively near. In the case of those emigrating from several Asian So¬
viet Republics, notably Uzbekistan and Kazakstan, and Romania, the
separation from a German State went back several centuries, and for
many years before the late 1980s there had been no exchange with, or
support from, a German-speaking community. Many Aussiedler, com¬
ing as they frequently do from isolated rural backgrounds, find
present-day West German values and life-styles difficult to understand.
With respect to the language, they may be unfamiliar with the norms
of usage of both Standard German as required in formal situations and
the colloquial forms appropriate for everyday conversation. In their
countries of origin they would have used Polish, Russian or Romanian
in formal contexts, and either a local variety of these languages or their
regional German dialect for more colloquial purposes.
14.1.3 New minorities

The ‘new minorities’ are the people who moved to west Germany from
the late 1950s onwards. In numerical terms they are more significant
tharK.the old ones, and the social, educational and linguistic issues that
they raise represent a real challenge to German society. As in other
European countries, the term ‘minorities’ is employed here in a
broader sense than, for instance, when we talk about ‘the Frisian mi¬
nority’ or ‘the Danish minority’ in the north of Germany. In the
present context the word is generally used in the plural, as it refers to
groups of people from a number of different countries of origin. Indi¬
vidual nationalities among the new minorities are referred to as ‘the
Turks’, ‘the Yugoslavs’, ‘the Italians’ etc., rather than Kurds, Croats,
Serbs, Macedonians and so on. In other words, their ethnic identity is
usually glossed over.
We can divide these new minorities into two groups:
(1) The migrant workers and their families (formerly referred to as

Gastarbeiter, ‘guestworkers’) account for some four million people.
They have been in the Federal Republic since the 1960s and have
MIGRATION 293
become, it seems, a permanent feature of German society. Later sec¬

tions of this chapter will be concerned mainly with this group.
(2) Political refugees, mainly from the southern parts of South-East

Asia, who entered the Republic in large numbers in the early 1980s.
Residence of these Asylanten (‘asylum seekers’) in the country is sub¬
ject to certain regulations and procedures, which in practice make it
difficult for them to have any form of close contact with the indigen¬
ous population. For instance, they are prevented from taking up
employment during their first years of stay in the Republic, and they
normally live in communal housing outside towns. They therefore have
little opportunity or motivation to learn German, at least until they
have been officially granted asylum, which many are not.
Government statistics used to list these two groups separately, but
now they refer to both of them with the collective term Ausldnder,
‘foreigners’/‘aliens’. By the end of 1986 there were Just over four and
half million ‘foreign nationals’ living in the Federal Republic, equal to
7.4 per cent of the total population of 61 million.
14.2 Migration
14.2.1 Background to migration to Germany

Germany has experienced large-scale immigration before. In fact,
every time its economy expanded significantly it became necessary to
recruit workers from abroad to meet new production requirements.
Thus, a census carried out in 1907, for instance, showed that nearly
800,000 foreign workers were employed in German industry and agri¬
culture; by 1914 the figure had risen to 1.2 million (Geiselberger
1972).
The latest big wave of migration this century started in 1955. By
that time the west German economy had readjusted itself to post-war
political conditions, had already absorbed the large number of refu¬
gees, and was still expanding. The west German Government, in
common with other European states, signed agreements with a number
of countries (e.g. Italy 1955, Spain and Greece 1960, Turkey 1961,
Portugal 1964, Yugoslavia 1968) to organize the recruitment of un¬
skilled labour. Private employers, too, launched recmitment drives,
with the result that the number of foreign workers in the Federal Re¬
public rose steadily: whereas in 1954 only 0.4 per cent of the total
REPUBLIC OF GERMANY
workforce consisted of foreigners (72,096 people), by 1964 the figure

had risen to 4.4 per cent (or 985,616). It reached its peak at 10.9 per
cent (or 2,286,625 workers) in 1974. Recruitment and continued em¬
ployment of foreign labour were dependent on the fluctuations of the
German economy, and the foreign workers thus served as a kind of
buffer (Konjunkturpuffer was the German term used). But as more and
more migrants brought their families to join them, planning a longer
stay in the Federal Republic, it became less and less possible to man¬
ipulate the number of foreigners.
When the west German economy started to suffer the effects of the
1970s oil crisis, recruitment of foreign workers was officially halted.
Government figures for 1987 suggest that there are at present some
1,900,000 non-German workers in the country, accounting for approxi¬
mately 9 per cent of the total workforce.
There are two significant features that distinguish the last of the
three migration periods this century from the earlier two. First, a large
number of those who went to work in the Federal Republic soon
brought their families as well, which in effect doubled the number of
foreign nationals living in the country. Nowadays the total number of
non-Germans registered is about 4.5 million (1986: 4,512,700) or some
7.4 per cent of the total population. This figure includes migrant wor¬
kers and their families, as well as immigrants granted political asylum
(1987 Datenreport).
Secondly, a large number of the migrants who originally went for a
limited period of time have stayed on. They represent migration in one
direction only, even if the possibility of returning, either voluntarily or
as a result of a government decision, always exists. The Federal Re¬
public already has, therefore, what amounts to a de facto first- and
second-generation immigrant population, which has a higher birth rate
than that of indigenous Germans. Official statistics forecast that, in
spite of a general decline in population from 61 million to 58 million
by the year 2000, the number of non-German residents is likely to
remain at its present level (4.5 million). This means that the proportion
of the total will increase to nearer 8 per cent.
MIGRATION 295
14.2.2 Composition of the new minorities and geographical

distribution
TABLE 14.1 Origins of West German foreign nationals
Country of origin Number of immigrants Percentage
Turkey 1,434,000 31.8

Yugoslavia 591,000 13.1
Italy 537,000 11.9
Greece 279,000 6.2
Spain 150,000 3.3
Source: Datenreport 1987
s compared to the 1970 figures, the total of Turkish nationals has

trebled, whereas the numbers for the other four groups have fluctuated
only slightly.
Foreigners have settled throughout the Federal Republic, but there
are marked discrepancies in patterns of settlement. Over 80 per cent of
the foreign population live in urban areas, particularly the large, heav¬
ily industrialized conurbations. 51 per cent live in cities of over
100,000 inhabitants (the proportion for the west German population as
a whole, by comparison, is about 33 per cent).
The cities with the highest numbers of foreigners are the following:
TABLE 14.2 Cities with the highest number of foreigners
City Number of foreigners
Berlin (West) 257,800

Munich 221,200
Hamburg 175,800
Frankfurt 148,300
Cologne 137,000
Stuttgart 102,600
REPUBLIC OF GERMANY
The cities with the highest proportion of foreigners are slightly differ¬
ent:
TABLE 14.3 Cities with the highest proportion of foreigners
City Proportion of foreigners (%)
Frankfurt 25.0
Offenbach 21.7
Stuttgart 18.2
Munich 17.4
Diisseldorf 16.1
Mannheim 15.5
Cologne 15.1
Ludwigshafen 14.3
Remscheid 13.9’
Berlin (West) 13.8
Heilbronn 13.1
Duisburg 13.0
Nuremberg 12.8
Source: Datenreport 1987
In the earlier years of recruitment, immigrants (mainly single men

or, 40 a lesser extent, single women) used to be housed in communal
quarters. This changed as migrants were joined by their dependants or
started new families in the Federal Republic. The great majority today
live in rented accommodation, more often than not in ghetto-like con¬
ditions in mn-down inner-city areas. Foreign nationals tend to be
concentrated in specific districts. In Berlin-Kreuzberg, for instance,
they form some 30 per cent of inhabitants, in Munchen-Ludwigsvor-
stadt 40 per cent, and around the main railway station in Frankfurt as
much as 80 per cent. Because Turkish nationals make up such a high
proportion of the total foreign population, they are particularly affected
by ghettoization.
Living under ghetto-like conditions breeds a host of social problems
and renders integration into West German society virtually impossible.
But it does foster a sense of cohesiveness among the minority, rein¬
forcing their sense of identity and contributing to the maintenance of
their languages (this point is taken up again in section 14.4).
MIGRATION 297
14.2.3 Reasons for migration and problems of remigration

Some foreign workers from Spain, Greece and Turkey migrated for
political reasons. For the majority, however, motivation was purely
existential - they came in search of work and in the hope of being able
to send money to their families back home and/or to save up to buy
land/set up business/build themselves a house when they returned.
Maria Borris (1973), in a case study of foreign workers in Frankfurt,
found that most of those she interviewed were long-term unemployed
in their country of origin: for example, 56 per cent of Spanish migrants
in Europe came from Galicia and Andalusia, traditionally the areas
with the highest unemployment rates in Spain. The reasons for migrat¬
ing given to her were always the same: the need to earn money to
survive, coupled with hopes for a better future on returning home. She
found that even recent arrivals, whom she would have expected to
have heard reports from others about the realities of migration, showed
an almost total lack of appreciation of working life in the Federal Re¬
public. More disturbing, she thought, was their rose-tinted view of their
own future prospects, as it is still the case today that for many migrants
returning to the home country becomes an economically sound prop¬
osition only when they reach retirement age.
Many thousands of migrants have returned to their own countries
upon completing their contracts of employment. But a comparison of
figures for those entering and those leaving the Federal Republic
shows a consistent surplus of entries, even after the Federal Govern¬
ment stopped recruitment, tightened restrictions on both entry and
residence, and launched a campaign to encourage remigration. The
longer people stay, the more secure their legal status becomes (i.e. they
are more likely to have their permits renewed) and the more stable
their economic situation is likely to be. As a consequence, they are less
motivated to return to their countries of origin.
Data from 1982 show the considerable proportion of migrants who
had been resident in the Republic for over six years. The figures con¬
firm the impression, gained in other countries as well, that these ‘new
minorities’ are developing into a permanent feature of modem Euro¬
pean society.
REPUBLIC OF GERMANY
Proportion of migrants from each country who have stayed in West

TABLE 14.4
Germany for six years or longer:
Country Length of stay

from six to over ten
eight years (%) years (%)
Spain 15.7 73.6

Greece 15.8 69.1
Yugoslavia 20.5 61.4
Italy 15 55.9
Turkey 27.8 36.3
Source: Informationen zur politischen Bildung, 201, 1984
Remigration, more often than not, is fraught with social, psycho¬

logical and practical problems which become worse the longer the
returning migrants have been away. Those who have not been able to
save up enough to ensure a stable income upon returning, and those
who face unemployment, often find themselves as much on the mar¬
gins of society as they were when they left. Their plight is sometimes
made worse by feelings of insecurity and a lack of identity caused by
their long absence. Their children, who may never have experienced
life, in their parents’ country, can be particularly hard hit. For instance,
it has been reported that young Turkish girls find that their new society
expects standards of behaviour, both at school and in general, with
which they are unfamiliar (Fritsche 1984; Berber 1984). In many cases
their knowledge of German will prove to be of little value to them in
their new lives, and through lack of practice they will gradually lose it
(Wolff 1984). In view of the serious problems that can arise, it is not
surprising that relatively few families have decided to remigrate volun¬
tarily.
14.2.4 Official policies towards migrants

For years the German Federal Republic did not have an official policy
towards foreigners (Ausldnderpolitik), apart from the recruitment of la¬
bour. Politicians and trade unionists used to insist that the Republic
was not a country of immigration, thus disclaiming responsibility for
MIGRATION 299
the foreign population or for the lack of special provision for it. While
the ‘economic miracle’ lasted, workers were hired on fixed-term con¬
tracts, and those who did not want to return when their contracts
expired usually had little difficulty in finding new employment. When,
in 1973, the oil crisis brought about a general recession, all recruitment
officially ceased, and measures were taken to restrict the issuing of
entry visas; more stringent conditions were set on the renewal of work
and residence permits. In 1975 Child Benefit regulations were changed
so as to allow payment, not according to the number of children in the
family, but only for those children living with their parents in the Re¬
public. However, migrants were still able to bring their families
(including all dependent relatives), in accordance with the policy of
Familiennachzug (‘family reunification’). Large numbers of depend¬
ants were, in fact, brought to the Federal Republic in the mid- and late
1970s (which for older children and adolescents, in particular, often
had a destabilizing effect). To avoid a situation in which the Familien¬
nachzug policy would result in more young people on the look-out for
employment being brought into west Germany, legislation was passed
in 1981 barring immigrants’ children of over fifteen from joining their
families; another law of the same year specified that, in general, a
four-year period between entry into the Federal Republic and the issue
of a work permit would apply.
Today less than 40 per cent of foreigners living in west Germany
form part of the working population. The rest are non-working women,
children, young people and unemployed.
The measures taken after 1973 halted the entry of new workers, but
they did not result in a significant reduction in the number of foreign
nationals resident in the country, as few of them opted to go back. The
economic crisis had hit other parts of the world as well, so there was
little point in leaving one precarious existence for another. The Federal
Government’s attempt to reduce the number of foreigners by promot¬
ing remigration (by, in effect, buying it) was only partly successful, as
the financial incentives were combined with severe conditions such as
a commitment never to return to west Germany to take up employment
again.
Official policies towards foreigners were designed to protect the
German labour market. This they have done, but at the same time they
have contributed to a rise in unemployment among migrants. Foreign¬
ers do not enjoy the same rights as German citizens living in the
Republic, and their permanence in the Federal Republic is uncertain.
REPUBLIC OF GERMANY
since it is subject to residence permits being renewed or revoked at the

discretion of the authorities. Yet the Government recognizes that the
German economy cannot do without its foreigners, since they (as a part
of the workforce, as consumers, taxpayers and ratepayers, and also as
employers, mainly in the hotel, catering and food retail industries) are
an integral part of West German economic life.
14.3 Sociolinguistic aspects
14.3.1 Socio-demographic composition

As mentioned before, the new minorities are composed of a variety of
nationalities, each with its own set of homogeneous linguistic, cultural
and religious characteristics. Germans may look upon them as one
group only (‘the foreigners’), but they themselves feel little sense of
cohesion between nationalities. Identification occurs only among mem¬
bers of the same group, who tend to define themselves in ethnic terms
(e.g. Kurds, Croats, Sicilians, etc.) rather than national ones. There can
even be animosity between the different groups (as is the case with
Turkish ones) who, depending on the region from which they originate,
may speak another language, have a different religion, belong to a dis¬
tinct culture, and differ in political views, and so reflect the same
rivalries that exist within Turkey.
Many of those who came to the Federal Republic were among the
least privileged members of their society in terms of employment op¬
portunities, housing and education. A good number of them exchanged
their previous positions for similar ones in the host country. They may
be financially better off than before, but their social status, level of
education and training and state of health tend to be low. As unskilled
workers they are more vulnerable than others to the ups and downs of
economic cycles, i.e. more likely to become unemployed than German
workers. As they have to move to wherever they can find jobs, they
have fewer opportunities for settling into a stable community.
Migrants not only form a ‘new’ minority. They are also, in a literal
sense, a young one. In 1982 one-third of all foreigners were under
twenty-one. The birth rate is highest among Turkish families, who ac¬
count for half the total of new births among migrants. In Italian, Greek
and Spanish families the birth rate is about the same as in German
families.
Second-generation immigrants have added new features to charac-
SOCIOLINGUISTIC ASPECTS 301
teristics of new minorities. By ethnic origin and nationality they are

non-German, although they may never have been to the country whose
nationals they are; they are bilingual, in many cases with German as
their dominant language; they are also likely to be bicultural. They are
therefore, at least potentially, intermediaries between the two cultures,
but this role is seldom appreciated by either side - the Germans con¬
sider them foreigners, and their own ethnic community may resent
their degree of accommodation to German customs and values. They
find themselves under considerable pressure from both sides to con-
fonn. They may feel that remaining ‘Turkish’, for instance, will please
their families and reinforce their sense of identity. It also makes remi¬
gration, which is always a possibility, less painful. Germanization, on
the other hand, can cause alienation from the family group, but at the
same time facilitate integration into west German society.
14.3.2 The migrants’ linguistic background

The majority of migrant workers moved to Germany from regions on
the fringes of their countries, in both geographical and economic
terms: Galicia and Andalusia in Spain, Sicily and the Mezzogiomo in
Italy, Macedonia in Yugoslavia and Greece, and Anatolia in Turkey.
The linguistic varieties spoken in these areas tend to be quite different
from the standard language. For this reason, and also because many
migrants have undergone only a minimum of schooling, their language
competence is mainly oral, and predominantly in the regional variety,
with varying degrees of fluency in the standard. In the Federal Repub¬
lic they find themselves in situations in which knowledge of the
standard form of their own language is required of them, as the auth¬
orities tend to employ interpreters and translators who use this variety.
A further, and possibly more important, problem is posed by the fact
that in west Germany, as in other western societies, there is a good
deal of reliance on the written mode for negotiations, agreements and
official communication in general. The syntax and lexicon of the writ¬
ten norm tends of course to be more complex or difficult than that used
in colloquial styles.
Lack of familiarity with the written form of their own language can
also prove to be disadvantageous to migrants. Illiteracy rates among
migrants are estimated at about 10 per cent (Auemheimer 1984), but
there are indications that it may be proportionately much higher among
women, since men will often have had a second chance to learn to read
REPUBLIC OF GERMANY
and write during their military service. Migrant women tend thus to
become much more linguistically isolated and more dependent on their
menfolk once they are separated from the wider family circle at home.
14.3.3 Problems associated with migrants’ learning of German

While there is little available research on migrants’ competence in, and
use of, their home language, a good deal has been published about
their acquisition of German and some aspects of their use of this lan¬
guage. Most migrants knew no German at all when they entered the
country, whether as workers or as dependants. In the early days, when
large numbers were contracted by particular firms, there were always
middlemen who could act as interpreters; later newcomers were helped
by fellow countrymen already settled in the country. The factors gov¬
erning the extent to which migrants become proficient in German are
numerous. In general, there is more pressure to learn German on those
who go out to work than there is on people who stay at home. In fact,
those who live in a ghetto-like environment have little need to learn
the language of the society at large. For the formal occasions when
contact with school, the authorities, health services, etc. is necessary,
children or friends can help to bridge the language gap.
The opinion that ignorance of the language of the host country is a
handicap, whatever the circumstances, is widespread. In most people’s
perceptions (Germans’ as well as migrants’) competence in German is
a prerequisite for successful integration into German society, which is
seen by many as the only satisfactory solution for the difficulties fac¬
ing migrants.
From 1971 to 1981 a major project was carried out in Wuppertal,
and later Hamburg and Kiel, to study the acquisition of German by
Spanish, Italian and Portuguese workers (Clahsen et al. 1983). Al¬
though language difficulties were usually stated as the main problem
experienced by the informants, the authors question the assumption
that a better knowledge of German would by itself speed up their inte¬
gration, arguing that an insufficient knowledge of the language
shielded migrants, at least to some extent, from a full appreciation of
negative attitudes on the part of Germans and their refusal to seek out
contact with them. Others (e.g. the sociologist Nikolinakos 1973) argue
that migrants are already integrated in that they have a place within
German industry as a kind of ‘reserve labour force’, mobile and avail¬
able when and where the need arises. In other words, their position
within the German social system is thought by some to be determined

by purely economic factors, and improved knowledge of German
would not alter this. Taking this point to its logical conclusion, Nikoli-
nakos argues that linguistic problems would disappear if the
socio-economic conditions for migrants were radically improved.
The issue can be approached differently, pursuing a socio-psycho-
logical, rather than economic, line of argument. Studies of migrants’
second language acquisition (some of which are listed below) consist¬
ently show that the most influential factors in successful acquisition are
motivation (to learn the language) and positive attitudes (towards lan¬
guage learning and towards the host society). Both of these factors
depend on the social environment in which the migrant finds him¬
self/herself. The linguist Schumann (1976) uses the concept of the
‘psychological and social distance’ that the learner perceives between
himself and the host country. He makes the point that the greater the
‘distance’ the more likely it is that acquisition will be hindered. By
gradual integration, however, this distance can be reduced or removed,
and language acquisition facilitated. In other words, integration and
language proficiency are interdependent. In this view, integration is a
contributory factor in the language learning process; it cannot be post¬
poned until some future time. This goes against the attitude, often
implicit in regulations or encountered among the public, to the effect
that migrants ought to demonstrate their mastery of a language before
integration can be contemplated. For instance, current legislation
concerning foreigners in the German Federal Republic (the Ausldn-
dergesetz) stipulates that a foreigner can apply for removal of residence
restrictions after five years’ residence in the country; the precondition
for obtaining it is that the applicant has become ‘integrated into the
economic and social life of the Federal Republic’, which means that
(s)he must prove that (s)he has a job, somewhere to live and oral
fluency in German. In effect, a certain degree of integration is required
of migrants even before they know whether they will be allowed to
stay in Germany. It is of course precisely during their first years of
residence in the country that they will have experienced the hardship
and the problems which affect their attitudes towards Germans and
West Germany negatively, thus lowering their motivation to learn the
language. Integration always requires considerable effort; and it seems
unrealistic to expect it from those who have not entered the country as
genuine immigrants.
The problems relating to integration of migrants and their uncertain
304 CASE STUDY III; MIGRANT WORKERS IN THE FEDERAL
REPUBLIC OF GERMANY
future represent two major obstacles to the acquisition of German.

There are additional reasons why so many never learn to master the
language of the host country, which are often of a practical nature. For
instance, few foreigners attend German classes. Meisel (1975) found
that only 12 per cent of the foreign workers he interviewed went to
such courses. The reasons they gave were tiredness after a long day’s
work, shift work, looking after children or lack of confidence in their
own ability to learn. Researchers claim that what German language
migrants know they have learned at work (Meisel 1975; Clyne 1977;
Klein and Dittmar 1979; Clahsen et al. 1983); but the amount of lin¬
guistic knowledge required there is limited, as are, in general, the
opportunities for communication with Germans at the workplace.
14.3.4 Factors affecting migrants’ acquisition of German

The continued presence of large numbers of foreigners with different
linguistic backgrounds having to communicate in German has
presented linguists with ideal conditions to undertake research into dif¬
ferent areas of contact linguistics, contributing to the rapid growth of
sociolinguistics and psycholinguistics as disciplines in west Germany.
In the 1970s a number of large-scale projects involving several dif¬
ferent national groups of migrants were carried out in various parts of
the country. These investigations were concerned with (1) the migrants’
acquisition of German, in an untutored and unstmcmred way, as dis¬
tinct. from formal learning by means of tuition; and (2) the features that
their German tended to show and the question of whether the so-called
Gastarbeiterdeutsch (German used by migrants) could be described as
a kind of pidgin or pidginized variety of the language.^
Since the end of the 1970s research interest has shifted towards the
second generation of migrants. Studies have been undertaken concern¬
ing general issues related to migrants’ children’s acquisition of German
and their maintenance or loss of the mother tongue. Other researchers
have focused their attention on specific linguistic features of their first
and second language, such as interference phenomena, code-switching
and communication strategies. A third group is interested mainly in the
education of migrant children, their integration into the German school
system (or lack of it), and the provision for the teaching of both the
mother tongue and German as a second language.^
In much of the research into (first generation) migrants’ acquisition
of German, the scholars concerned were keen on isolating factors that
seemed to facilitate the acquisition process. Klein and Dittmar (1979)

worked with a sample of forty-eight Italian and Spanish adult inform¬
ants, two-thirds men and one-third women, and they used both
interviewing and participant observation methods. They decided to
look at nine factors which they thought were relevant, making a dis¬
tinction between (a) ‘bias factors’, of a personal and individual type,
and (b) ‘environmental factors’, which referred largely to the social
context.
(a) Bias factors:

(1) origin;
(2) age (at time of migration);
(3) attendance at school;
(4) formal/professional qualifications;
(5) sex.
(b) Environmental factors:

(1) contact with Germans at the workplace (opportunities and fre¬
quency);
(2) contact with the German language during leisure time (type and
frequency);
(3) type of accommodation;
(4) duration of stay, i.e. how long they had been in Germany.
The informants were subdivided into three groups, according to

their ‘syntactic level’ and the scores for the variables (the above fac¬
tors) were then correlated with each informant’s syntactic level.
‘Contact with Germans during leisure time’ turned out to be the most
significant factor, followed by ‘age’ and ‘contact at workplace’; ‘ac¬
commodation’ was relevant when it co-varied with ‘contact with
Germans’, and ‘duration of stay’ seemed to be important only for the
first two years. The authors were surprised to see the variable ‘age’ as
having the second strongest correlation with syntactic achievement,
given that only adults were interviewed. They concluded that ‘age’
may well co-vary with environmental factors, a young person being
more willing and able to build up social contacts with Germans than an
older one. Orlovic-Schwarzwald (1978) and Keim (1984), too, found
‘age’ to be an important factor. Keim argues that older learners can
become equally proficient in two languages as younger ones, but the
conditions under which successful acquisition takes place may be dif¬
ferent - if social and psychological barriers are removed they no
REPUBLIC OF GERMANY
longer hinder language acquisition. Apart from ‘contact with Germans’

(in leisure time and at work), Keim (1984) also found that three other
factors were beneficial to the acquisition of German by her informants,
who were twelve adult Turkish migrants:
(1) the firm intention, at the time of migration, to stay in the Federal
Republic for a long time;
(2) a critical and reserved attitude towards current political and relig¬
ious tendencies in Turkey (which made return less attractive);
(3) an appreciation of the legal situation in Germany vis-d-vis their
status as foreigners.
These three are called ‘internal factors’ by Keim, as they reflect

emotional ties, experience and motivation. They are, in her view, more
important to second language acquisition than ‘external factors’, such
as age, accommodation and qualifications, the possibly negative effects
of which are surmountable.
It is perhaps interesting to see the relative insignificance of the vari¬
able ‘duration of stay’, as many researchers (e.g. Meisel 1975; Klein
and Dittmar 1979; Keim 1984) had originally thought that it was more
important. In all studies, informants who had been in Germany for
more than ten years varied in their ‘syntactic level’. Some researchers
have examined the possible reasons why the German of some migrants
never progresses beyond a certain level. Klein and Dittmar suggested
that:
X.
... it may well be that during the first two years a certain syntactical level
is gradually acquired whose height depends on other factors such as ‘con¬
tact’, ‘age’ etc. The acquisition process then slows down, or even stops,
and may well start moving again when the environmental factors change.
(Klein and Dittmar 1979: 207).
Environmental as well as psychological factors, then, can be seen to

have a bearing on second language acquisition. They will often co-vary
and cause a chain reaction, either increasing or reducing the pressure
and/or opportunity to acquire German. It is clear, however, that few of
these factors come within the control of the migrants themselves, and
even if societal attitudes and material conditions changed dramatically
in favour of the migrants, their successful acquisition of German would
not thereby be secured automatically.
THE EDUCATION OF MIGRANT WORKERS’ CHILDREN 307
14.4 The education of migrant workers’ children
14.4.1 The EC Directive on the education of migrant workers’

children
At an early stage of migration the Council of Europe began to show
concern about the education of migrant workers’ children. Several res¬
olutions were passed demanding that migrant workers and their
families should be taught the language of the host country, that
mother-tongue teaching should be made available, that migrant child¬
ren should be supported within the national school systems, and that
provision should be made for the smooth reintegration of those child¬
ren whose parents returned to their countries of origin. In the 1970s the
Standing Conference of European Ministers of Education began to turn
their attention to these problems, along with other international organ¬
izations, such as UNESCO, OECD and the International Labour
Office. The general consensus that emerged can be summarized as fol¬
lows:
(1) migrant children should enjoy the same educational opportunities
as children from the host country;
(2) segregation should be avoided by encouraging transfer from pre¬
paratory clases to mainstream education as early as possible and
by stimulating and promoting international understanding among
indigenous children;
(3) children should be able to learn, or continue to develop, their
mother tongue at school.
In 1974 the EC authorities initiated a legislative process which re¬

sulted in 1977 in the adoption of Council Directive 77/486/EEC on the
education of the children of migrant workers. As a directive, it is
legally binding and it applies to children of migrant workers from EC
member states as well as non-member ones (which is particularly im¬
portant in the British context). The directive originally proposed by the
Commission was more progressive in outlook and went further in its
detailed provisions. Its terms, however, were toned down in certain
respects by the Council, as some of its members (e.g. Britain and the
Federal Republic of Germany) favoured a less binding form of words.
This affected, in particular. Article 3, which deals with the provision of
mother-tongue teaching. It was as a result of this EC directive that
educational provision for migrants’ children became part of general
educational planning.
REPUBLIC OF GERMANY
The directive requires member states to make free tuition available

to the children of migrant workers in the language of the host country
and also in the children’s mother tongue. However, it allows member
states to do so ‘in accordance with their national circumstances and
legal systems’ (a phrase which is repeated in Articles 2, 3 and 4). This
means, in effect, that member states have a great deal of freedom in
deciding the amount and kind of resources they are willing to devote to
such programmes. Educational provision for migrant children has not,
in fact, proved to be high on the list of priorities of any member state.
(See Appendix A for the full text of the directive).
14.4.2 Educational provision for migrant workers’ children in the

Federal Republic
According to the German Constitution, education is the responsibility
of each individual Federal State or "Land’. Thus each Land is free to
decide on the detailed provision at all levels, including matters of re¬
source allocation, types of schooling and curriculum details. An
institutionalized link between the eleven Lander ministries of educa¬
tion is provided by the Standing Conference of Education Ministers.
This body formulates general policy and issues guidelines that the indi¬
vidual states are free to carry out in their own way. With respect to the
education of migrants’ children, the Standing Conference adopted in
1976 two principles which are still considered valid today. The first is
that the objective of the education of migrants’ children should be:
(1) to integrate these children into the German educational system;

(2) to protect them from ‘over-germanization’ through the develop¬
ment of their mother tongue and native culture, so as to enable
them to identify with the culture of their parents and to ease their
return to their parents’ homeland.
The second principle is that this double objective should be im¬

plemented in the following ways:
(1) the same compulsory school attendance should apply to foreign

children as to German ones (i.e. from seven to eighteen years of
age: full-time attendance for nine years, then part-time until
eighteen);
(2) special provision should be made for German language teaching
in the case of children whose command of German is insufficient
for them to follow mainstream education;
(3) migrants’ education should not separate them from German

children, i.e. they should be taught together;
(4) provision should be made for children to be taught their parents’
language.
The various Lander have developed different programmes with re¬

gard to the education of the children for whom they are responsible. A
combination of the following types of programmes has usually been
available:
(1) Preparatory classes (Vorbereitungsklassen) are intended to provide

intensive German language teaching for groups of about fifteen child¬
ren with the same or different language backgrounds. Attendance can
extend for up to two years, after which the children join mainstream
education. Apart from the specialist language tuition, these courses
should follow the same curriculum as those for German children. If the
children in the group have the same mother tongue instruction can take
place, part of the time, in this language and be in the charge of tea¬
chers who speak the same native language. Transfer to mainstream
education is supposed to take place as soon as the child has reached a
sufficient level of knowledge of German to enable him/her to follow
the normal curriculum.
(2) In bilingual classes German is taught as a foreign/second language

by German teachers and other subjects - some or all of them - by
foreign teachers with the same mother tongue as the pupils. Again, it is
envisaged that transfer to mainstream education will take place as early
as possible.
(3) Intensive courses {Forderkurse) are intended to offer flexible provi¬

sion of German language tuition to individual pupils or small groups
for varying lengths of time. These courses are offered during school
hours, perhaps during the periods when lessons deemed to be of a
more peripheral nature (domestic science, P.E., art or music) are taking
place.
(4) Mother-tongue classes are designed to teach the language, history,

geography and aspects of culture of the migrants’ home country. They
do not form part of the West German education system, and they are
held outside school hours. Some of them provide a mixture of lan¬
guage tuition and background knowledge, others attempt to help
children to obtain the school qualifications of the countries concerned.
REPUBLIC OF GERMANY
They are financed by the variojus migrant communities themselves or

by the respective national governments as in the cases of Spain and
Italy. They vary considerably as to the demands they make of children
- also as far as the teachers’ own preparation is concerned. In the case
of some countries, the respective ministries of education make resour¬
ces (teachers and/or materials) available. Religious institutions can also
be involved, as it is in the case of Islamic schools (Koranschulen).
Some Lander support these classes financially and by cooperating in
the design of teaching programmes; in return for this, they may assume
overall control of the scheme - but this does not apply to Islamic
schools, which concentrate on the teaching of the Muslim faith and of
the Arabic required for religious purposes, rather than the teaching of
Turkish.
14.4.3 Problems involved in the education of migrants’ children

Because of the uneven distribution of migrants in the Federal Republic
and the different measures taken by the various Lander, it is not
possible to give a global assessment of the educational measures out¬
lined above. Also, the instability of the immigrant population and the
sharp increase in the number of migrants’ children attending West Ger¬
man schools make it difficult to obtain a generalized picture. In some
areas the proportion of foreign children has risen to 30 per cent, 40 per
cent or even 50 per cent of the total school population within a decade,
thiis compounding the problems of provision, planning and resourcing.
Among those who have reported on the education of migrants’ children
there is general accord about the nature of the problems; and they all
agree that only massive new allocation of resources can solve them.
But there are other problems (e.g. those relating to non-attendance at
school and growing up against socially and culturally deprived back¬
grounds) which cannot be tackled simply by greater financial
provision; they usually affect the children of Turkish migrants, who far
outnumber those from other countries. Members of Italian, Spanish,
Greek (who now enjoy EC citizenship status) and Yugoslav migrant
groups tend to be more widely dispersed, and therefore proportionately
fewer of their children attend the same school. In addition, the majority
of them have been in the Federal Republic for a long time and have
become better adjusted and more integrated.
The issues causing concern are those posed by children bom outside
Germany who joined their parents in the Republic after starting school
in their home countries, the Seiten- or Spdteinsteiger (‘late joiners’).

When they enter German schools they have no knowledge of the lan¬
guage, and even after attending preparatory or intensive classes they
may never acquire enough to be able to follow mainstream education.
Because of their lower achievement levels, they are not normally
placed in the same classes as German children of the same age group,
and they are thus even less likely to make meaningful contact with
German children as, in addition to the language gap, there is an age
gap as well.
Children who were bom in the Federal Republic and attended kin¬
dergartens or playgroups are usually bilingual upon entering school,
and they experience little, if any, language difficulty. But many of
those who have had no contact with German-speaking children before
starting school are disadvantaged, since they have to acquire German
in preparatory classes before they are able to join regular ones.
The best results, in educational terms, seem to have been achieved
when a system of bilingual education has been followed. Different
models (nationale Modelle) have been devised. They are all still at an
experimental stage. Generally, they consist of the majority of classes
being taught in the children’s mother tongue; only German lessons and
(usually) those in art, music and RE. are conducted in German. Child¬
ren have been observed to progress well (Oomen-Welke 1985) if they
do not have to compete against German children - the West German
educational system is quite competitive.^ But bilingual education
means that foreign children are segregated from German ones, at least
for a part of their education, and thus this type of education pro¬
gramme contributes relatively little to the declared aim of integration at
least in the short term.
Critics have argued for some time that the various educational
measures taken in the Federal Republic have not made any significant
contribution to the improvement of educational progress among mi¬
grants’ children. Some groups have done better than others, but overall
statistics relating to foreign children in German educational institutions
appear to bear out the claim. In 1985 82 per cent of migrants’ children
left school after the compulsory nine years, some two-thirds of them
without any formal qualifications; the corresponding figures for Ger¬
man children were 35.5 per cent and 6.6 per cent. An equally
disproportionately large section of them are sent to schools for pupils
with special difficulties. It has also been suggested that an alarmingly
high number of foreign children do not attend school at all: estimates
REPUBLIC OF GERMANY
vary between 8 per cent and 25 per cent of the foreign school popula¬
tion (Auemheimer 1984). The employment prospects of these children
are poor, as they have to compete with German school-leavers in the
job market. It seems likely that the children will perpetuate the uncer¬
tain, low-status position their parents occupy within the economic and
social system in the host country.
Schools have been criticized for not playing a significant role in the
acquisition of German by migrant’s children. Kutsch and Desgranges
(1985) report that they saw little progress made in German by the
children they observed, adding that the language is acquired more ef¬
fectively outside school, as part of the general socialization process.
The schools are up against two major obstacles. The first is at¬
tributed to the migrants themselves, that is, to their special social and
legal position, and their heterogeneous cultural and linguistic back¬
grounds. The second is a matter of educational provision. Most West
German teachers are ill-equipped to cope with what at times can be
quite large numbers of migrants’children in their classes, children who
often have special linguistic and educational needs. Few teachers have
followed training courses in German as a second language, even fewer
know any of the languages of their pupils, and there is a scarcity of
suitable materials, which makes special courses more difficult to pre¬
pare. Lessons with mother-tongue teachers can be problematic if the
latter are unfamiliar with German and the German school system. Es¬
sentially, neither German-bom nor mother-tongue teachers can be
expected (without some form of retraining) to have any real under¬
standing of bilingualism and biculturalism.
It is clearly important that everybody involved in the upbringing
and education of migrants’ children recognizes that they have special
needs. They live in an environment where they are surrounded by two
languages and two cultures, and access to both should be considered to
be their right. It is only if bilingualism and biculturalism are estab¬
lished successfully that these children’s integration into German
society, or into that of their parents’ home countries should the family
remigrate, can be ensured.
Notes
1. Because of its concern with western Europe this chapter is con¬
cerned with that part of Germany only which was West Germany
before 3 October 1990. In the east of the new Germany there
NOTES 313
also exists a linguistic minority, the Sorbs, who use Serbian, a

Slavonic language.
2. For example, the 1975 HPD Project {Heidelberger Forschungs-
projekt Pidgin Deutsch) led to the publication of Klein and
Dittmar’s (1979) study of migrants’ developing grammars and
Klein’s (1984) work on second language acquisition. Meisel
(1975) and Schumann (1978) consider the question of pidginiza-
tion, as did Clyne (1977) in a review article. Clahsen, Meisel and
Pienemann (1983) report on the results of the ZISA Project,
which looked at the acquisition of German by Italian, Spanish
and Portuguese workers. Orlovic-Schwarzwald (1978), Fritsche
(1982) and Keim (1984) focus on Turkish migrants, and Stolting
(1975; 1980) on Yugoslav ones. Multilingualism within the fam¬
ily and also in the wider context of an urban area was the subject
of a project carried out by the Institut fiir deutsche Sprache, re¬
ported on in Bausch (1982).
3. Among those who have reported on more general issues of child¬
ren’s acquisition of German are Nelde et al. (1981), Pfaff (1981)
and Rath (1983). Pfaff (1980), Pfaff and Portz (1980) and Piene¬
mann (1980) concentrate on linguistic aspects. A major project,
large-scale and long-term, dealt with various aspects of Italian
migrants’ children’s use of both Italian (standard and dialect) and
German - see Auer (1980; 1981; 1982; 1984) and Auer and di
Luzio (1984). A long-term project at the University of Saar-
briicken concerned itself with communication by migrants and
covered also the acquisition of German by Turkish migrants’
children (Kutsch and Desgranges 1985). A good number of
studies of linguistic and educational issues have been carried out
at the Centre for Multilingualism of the University of Hamburg,
e.g. by Rehbein (1985; 1987).
4. Failure in school achievement carries with it the threat of either
having to repeat the year or, even worse, being sent to a special
school for children with behavioural or educational difficulties.
Appendix
Council of the European Communities Directive of 25 July 1977 on

the education of the children of migrant workers (77/486/EEC).
THE COUNCIL OF THE EUROPEAN COMMUNITIES,
Having regard to the Treaty establishing the European

Economic Community, and in particular Article 49 thereof.
Having regard to the proposal from the Commission,
Having regard to the opinion of the European Parliament,
Having regard to the opinion of the Economic and Social

Committee,
Whereas in its resolution of 21 January 1974 concerning a social ac¬

tion programme, the Council included in its priority actions those
designed to improve the conditions of freedom of movement for wor¬
kers,^ particularly with regard to the problem of the reception and
education of their children;
Whereas in order to permit the integration of such children into the

educational environment and the school system of the host State, they
should be able to receive suitable tuition including teaching of the lan¬
guage of the host State;
Whereas host Member States should also take, in conjunction with the
Member States of origin, appropriate measures to promote the teaching
of the mother tongue and of the culture of the country of origin to the
above-mentioned children, with a view principally of facilitating their
possible integration into the Member State of origin.
NOTES 315
HAS ADOPTED THIS DIRECTIVE:
Article I
This Directive shall apply to children for whom school attendance is
compulsory under the laws of the host State, who are dependants of
any worker who is a national of another Member State, where such
children are resident in the territory of the Member State in which that
national carries on or has carried on an activity as an employed person.
Article 2
Member States shall, in accordance with their national circumstances
and legal systems, take appropriate measures to ensure that free tuition
to facilitate initial reception is offered in their territory to the children
referred to in Article 1, including, in particular, the teaching - adapted
to the specific needs of such children - of the official language or one
of the official languages of the host State.
Member States shall take the measures necessary for the training
and further training of the teachers who are to provide this tuition.
Article 3
Member States shall, in accordance with their national circumstances
and legal systems, and in cooperation with States of origin, take appro¬
priate measures to promote, in coordination with normal education,
teaching of the mother tongue and culture of the country of origin for
the children referred to in Article 1.
Article 4
The Member States shall take the necessary measures to comply with
this directive within four years of its notification and shall forthwith
inform the Commission thereof.
The Member States shall also inform the Commission of all laws,
regulations and administrative or other provisions which they adopt in
the field governed by this Directive.
Article 5
The Member States shall forward to the Commission within five years
of the notification of this Directive, and subsequently at regular inter¬
vals at the request of the Commission, all relevant information to
enable the Commission to report to the Council on the application of
this Directive.
Article 6
This Directive is addressed to the Member States.
References
Note: This list contains the references mentioned in the individual

chapters, and also the titles included among the suggestions for further
reading.
abeyA, m (1985) Valoraciones del programa de inmersion en Santa Coloma de

Gramanet, in M Siguan, (ed): Ensenanza en dos lenguas y resultados esco-
lares
ADLER, M (1977) Collective and Individual Bilingualism: a sociolinguistic
study. Helmut Buske Verlag, Hamburg
AELLEN, c and LAMBERT w E (1969) Ethnic identification and personality adjust¬
ments of Canadian adolescents of mixed English-French parentage,
Canadian Journal of Behavioural Science 1, 69—86
AGUIRRE, A (1985) An experimental study of code alternation, International
Journal of the Sociology of Language 53, 59-82
ALBERT, M L and OBLER, L K (1978) The Bilingual Brain: neuropsychological
and neurolinguistic aspects of bilingualism. Academic Press, New York
ALLARDT, E (1979) Implications of the ethnic revival in modem industrialized
society: a comparative study of the linguistic minorities in western Europe,
Commentationes Scientiarum Socialum 12, Societas Scientiamm Fennica,
Helsinki
—(1984) What constitutes a language minority? Journal for Multilingual and
Multicultural Development 5, Nos 3 & 4, 195-205
AMASTAE, J and ELIAS-OLIVARES, L (eds) (1982) Spanish in the United States:
sociolinguistic aspects, Cambridge University Press, Cambridge
APPEL, R and MUYSKEN, p (1987) Language Contact and Bilingualism, Edward
Arnold, London
ARNAU, J and BOADA, H (1986) Languages and school in Catalonia, Journal of
Multilingual and Multicultural Development 7, 2 & 3, 107-122
REFERENCES 317
ARNBERG, L (1981) A longitudinal study of language development in four

young children exposed to English and Swedish in the home, Linkoping
Studies in Education, Report 6, Department of Education, University of
Linkoping
—(1987) Raising Children Bilingually: the pre-school years. Multilingual Mat¬
ters, Clevedon
ARNBERG, L and ARNBERG, P (1984) The relation between code differentiation
and language mixing in bilingual 3- to 4-year-old children, in Scandinavian
Working Papers in Bilingualism, 2, 43—62
ARSENIAN, S (1937) Bilingualism and Mental development: a study of the intel¬
ligence and the social background of bilingual children in New York,
Teachers’ College, Columbia University
AUER, J C P (1980, 1981, 1982) Projekt: Untersuchungen zur Muttersprache
italienischer Migrantenkinder, Papiere des Sonderforschungsbereichs, 44,
54, 71, 79, 99, Universitat Konstanz
—(1984) On the meaning of conversational code-switching, in AUER and di
LUCIO (eds) Interpretative Sociolinguistics, Gunter Narr, Tubingen, 87-
112
AUER, J c P and di LUCIO, A (1984) Interpretative Sociolinguistics, Gunter Narr,
Tubingen
AUERNHEIMER, G {19S4) Handwbrterbuch Ausldnderarbeit, Beltz, Basel
AZEVEDO, H M (1984) The reestablishment of Catalan as a language of culture,
Hispanic Linguistics, 1, 2, 305-30
BAETENS BEARDSMORE, H (1974) Development of the compound-coordinate dis¬
tinction in bilingualism. Lingua 33, 2, 123-27
—(1979) European Schools, in Mpdet mellem sprogene i det dansk-tyske
groensomrade, Abenr4 Institut for Graensregionsforskning, 38^9
—(1980) Bilingual Education in International schools, European schools and
Experimental schools: A comparative analysis, in Lim Kiat Boey (ed) Bi¬
lingual Education, Singapore University Press, Singapore, 3-19
—(1981) Elements of bilingual theory, in H Baetens Beardsmore (ed): 1981
Study Series of the Tijdschrift van de Vrije Universiteit Brussel
—(1982) Bilingualism: Basic Principles, Multilingual Matters, Clevedon
—(1983) Substratum, adstratum and residual bilingualism in Brussels, Journal
for Multilingual and Multicultural Development 4(1), 1-15
BAIN, B (1976) Verbal regulation of cognitive processes: a replication of Luria’s
procedures with bilingual and multilingual infants. Child Development 47,
543-46
BAKER, C {19^5) Aspects of bilingualism in Wales, Multilingual Matters, Cleve¬
don
—(1988) Key Issues in Bilingualism and Bilingual Education, Multilingual
Matters, Clevedon
BAUSCH, K (1981) Mehrsprachigkeit in der Stadtregion. Jahrbuch 1981 des In-
stituts fiir deutsche Sprache, Padagogischer Verlag Schwann
318 REFERENCES
BEN-ZEEV, s (1976) What strategies do bilinguals use in preventing their lan¬

guages from interfering with each other, and what effect does this have on
non-language tasks? Papers and Reports on Child Language Development
12, Department of Linguistics, Stanford University
—(1977) Mechanisms by which childhood bilingualism affects understanding
of language and-cognitive structures, in P Hornby, (ed) Bilingualism, Aca¬
demic Press, London 29-55
BERBER, I (1984) Soziale und sprachliche Probleme der Remigration tiirkischer
Jugendlicher, Informationen Deutsch als Fremdsprache, 3, Jahrgang 84/85,
August 1984, DAAD, 37-41
BERGMAN, c (1976) Interference vs. independent development in infant biling¬
ualism, in G Keller, et al. (ed) Bilingualism in the Bicentennial and Beyond,
Bilingual Press, Editorial Bilingue, New York, 86-95
BERK-SELIGSON, s (1986) Linguistic constraints on intrasentential code-switch¬
ing: a study of Spanish-Hebrew bilingualism. Language in Society 15,
313-48
BERLIN, c S et al. (1972) Dichotic right ear advantage in males and females aged 5
to 10, Journal of the Acoustical Society of America 53, 368 (Abstract)
BERMAN, R A (1979) The re-emergence of a bilingual: a case study of a He-
brew-English speaking child. Working Papers on Bilingualism 19,
November 1979, Ontario Institute for Studies in Education, 157-79
BERRUTO, G and BURGER, H (1987) Aspekte des Sprachkontaktes Italienisch-
Deutsch im Tessin, Linguistische Berichte III, Westdeutscher Verlag,
Opladen 367-80
BLOOMFIELD, L (1927) Literate and illiterate speech, American Speech, 2, 423-
39 (reprinted in Hymes 1964), Language in Culture and Society, Row and
Harpers, New York
—(1933) Language, Holt Rinehart and Winston, New York (published in 1935
b^ George Allen and Unwin, London)
BLOM, J p and GUMPERZ, j (1972) Social meaning in linguistic structure: code¬
switching in Norway, in J Gumperz and P Hymes, (eds) Directions in
Sociolinguistics, Basil Blackwell, Oxford, 407-34
BOOS-NUNNING, K et al. (1986) Towards Intercultural Education, GILT (Alfa)
BORRis, M (1973) Ausldndische Arbeiter in einer Grofistadt, eine empirische
Untersuchung am Beispiel Frankfurt, Europaische Verlagsanstalt, Frankfurt
BRAINERD, c J and PRESSLEY, M (eds) Progress in Cognitive Development Re¬
search: verbal processes in children. Springer, New York
BREITENBACH, B von (1982) Italiener und Spanier als Arbeitnehmer in der BRD,
eine vergleichende Untersuchung zur europdischen Arbeitsmigration, Kaiser
Griinewald, Munich and Mainz
BROWN, R (1973) A First Language, George Allen and Unwin, London
BRUNER, J (1975) The ontogenesis of speech acts. Journal of Child Language
2, 1-19
REFERENCES 319
Bundeszentrale fur Politische Bildung, Stadstisches Bundesamt (Hrsg), Daten-

report 1987, Zahlen und Fakten iiber die Bundesrepublik Deutschland,
Band 257, Bonn
BURLING, R (1959) Language development of a Garo- and English-speaking
child. Word 15, 48-68
BYRAM, M (1986a) Schools in ethnolinguistic minorities. Journal for Multilin¬
gual and Multicultural Development, 7, 2 & 3, 97-106
—(1986b) Minority Education and Ethnic survival. Multilingual Matters,
Clevedon
—(1986c) Cultural studies in modern language teaching. Language Teaching
19, 322-36
—(1988) Cultural Studies in Eoreign Language Education, Multilingual Mat¬
ters, Clevedon
BYRAM, M and LEMAN, J (1989) Bicultural and trilingual education: the Foyer
Model in Brussels, Multilingual Matters, Clevedon
CALSAMIGLIA, H and TUSON, A (1984) Use of languages and code-switching in
groups of youth in a ‘barri’ of Barcelona: communicative norms in sponta¬
neous speech. International Journal of the Sociology of Language 47,
105-21
CARR, R (1973) Spain: 1808-1936, Oxford University Press, Oxford
CARROW, E (1971) Comprehension of English and Spanish by pre-school Mexi-
can-American children. Modern Language Journal 55, 299-307
CELCE-MURCIA, M (1978) The simultaneous acquisition of English and Erench
in a two-year-old child, in E Hatch Second Language Acquisition: a book of
readings, Newbury House, Rowley
CELLARD, J (1976) Alsacien: la langue du foyer. Le monde de I’ education (Sep-
tembre), 16-18
CHILD, I L(1943) Italian or American? The second generation in conflict, Rus¬
sell and Russell, New York
CHRISTOPHERSON, P (1948) Bilingualism, Methuen, London
CHURCHILL, s (1986) The Education of Linguistic and Cultural Minorities in
OECD Countries, Multilingual Matters, Clevedon
CLAHSEN, H, MEISEL, J AND PIENEMANN, H (1983) Deutsch als Zweitsprache:
Der Spracherwerh ausldndischer Arbeiter, Gunter Narr Verlag, Tubingen
CLYNE, M (1977) The speech of foreign workers in Germany (review article),
Language in Society 6, 268-74

_(1988) Budding linguist: some expressions of metalinguistic awareness in a
bilingual child; paper presented at the 1st Hamburg symposium on Multilin¬
gualism September 1988
COBARRUBIAS, J (1983) Ethical issues in status planning, in J Cobarrubias and
J Fishman (eds) Progress in Language Planning, Mouton, de Gmyter, Ber¬
lin, 41-86
COBARRUBIAS, J and FISHMAN, J (1983) Progress in Language Planning: Inter¬
national Perspectives, Mouton, de Gruyter, Berlin
320 REFERENCES
COLE, R (1975) Divergent and convergent attitudes towards the Alsatian dialect,
Anthropological Linguistics, 17:6
CORBELLA, J (1988) General survey of the 1980s: Social communication in
Catalonia, Generalitat de Catalunya, Barcelona
CORNEJO, R (1973) The acquisition of lexicon in the speech of bilingual child¬
ren, in P Turner (ed) Bilingualism in the Southwest, University of Arizona
Press, Tucson (revised edn, 1982), 141-69
CRYSTAL, D (1987) The Cambridge Encyclopedia of Language, Cambridge
University Press, Cambridge
CUMMINS, J (1976) The influence of bilingualism on cognitive growth: a syn¬
thesis of research findings and explanatory hypothesis. Working Papers on
Bilingualism, Ontario Institute for Studies in Education 1, 1-43
—(1978a) Metalinguistic development of children in bilingual education pro¬
grams: data from Irish and Canadian Ukranian—English programs, in: M
Paradis (ed) Aspects of Bilingualism
—(1978b) The cognitive development of children in immersion programs, Ca¬
nadian Modern Language Review 34, 855-83
(1978c) Bilingualism and the development of metalinguistic awareness,
Journal of Cross-Cultural Psychology 9, 131-49
—(1978d) Educational implications of mother-tongue maintenance in minority-
language groups, Canadian Modern Language Review 34 (3), 345^16
—(1979a) Cognitive/academic language proficiency and linguistic interdepend¬
ence, the Optimum Age question and some other matters. Working Papers
on Bilingualism 19, November 1979, Ontario Institute for Studies in Educa¬
tion, 197-205
—(1979b) Linguistic interdependence and the educational development of bil¬
ingual children, Review of Educational Research 49, 2, 222-51
—(1981) Bilingualism and Minority Language Children, Ontario Institute for
Sibdies in Education, OISE Press, Toronto
—(1984a) Bilingualism and Special Education: Issues in assessment and peda¬
gogy, Multilingual Matters, Clevedon
(1984b) Bilingualism and cognitive functioning, in S Shapston and D’Oyley
(eds) Bilingual and Multicultural Education: Canadian Perspectives, Multi¬
lingual Matters, Clevedon, 55-67
—(1984c) Linguistic minorities and multicultural policy in Canada, in J Ed¬
wards (ed) Linguistic Minorities, Policies and Pluralism, Academic Press
London, 81-105
CUNZE, B (1980) Lo sviluppo linguistico di una bambina bilingue italo-tedesca
(PhD thesis, University of Rome)
DARCY, N T (1953) A review of the literature on the effects of bilingualism upon
the measurement of intelligence, Journal of Genetic Psychology 82, 21-57
DATO, D p (ed) (1975) Georgetown University Round Table on Language and
Linguistics, Georgetown University Press, Washington DC
REFERENCES 321
DE AVILA, E (1987) Bilingualism, cognitive function and language minority

group membership, in P Homel, M Palij and J Aaronson, (eds) Childhood
Bilingualism, Lawrence Erlbaum Associates, New Jersey, 149-169
DE CICCO G and MAKING, J M (1983) Diglossia, regionalism and national lan¬
guage policy: a comparison of Spain and the Phillipines, in A Miracle, (ed):
Bilingualism, University of Georgia Press, 38-53
DE HOUWER, A (1990) The acquisition of two languages from birth: a case
study, Cambridge University Press, Cambridge
DE JONG, E (1986) The Bilingual Experience, Cambridge University Press,
Cambridge
DEMONTE, V and ORTEGA, s (eds) (1982) El bilingiiismo: problematica y reali¬
dad, Revista de Occidente 10-11, Extraordinario ft, Madrid
Departament de cultura de la Generalitat de Catalunya (1988) The Press in
Catalonia in the 1980s, Barcelona
DE VILLIERS, A and DE VILLIERS, J G (1979) Early Language, Eontana, London
DIEBOLD, A R (1961) Incipient bilingualism, Language xxxvu, 97-112
DILLER, K c (1972) ‘Compound’ and ‘coordinate’ bilingualism: a conceptual
artefact, Word 26, 2, 254-61
Dl LUZio, A (1984) On the meaning of language choice for the sociocultural
identity of bilingual immigrant children, in Auer and di Luzio (eds) Inter¬
pretative Sociolinguistics, Gunter Narr, Tubingen, 55-85
DODSON, c J (1981) A reappraisal of bilingual development and education:
some theoretical and practical considerations, in H Baetens Beardsmore (ed)
Elements of Bilingual Theory, Free University of Bmssels, 14—27
DORIAN, N (1981) Language Death: the Life Cycle of a Scottish Gaelic Dialect,
University of Pennsylvania Press, Philadelphia
DOYLE, A B, CHAMPAGNE, M and SEGALOWITZ, N (1978) Some issues in the as¬
sessment of the consequences of early bilingualism, in M Paradis, (ed)
Aspect of Bilingualism, Hornbeam Press, Columbia SC, 13-20
EASTMAN, c M (1983) Language Planning: an introduction. Chandler and
Sharp Publishers, San Francisco
EDWARDS, J (1977) ethnic identity and bilingual education, in H. Giles (ed)
Language, ethnicity and intergroup relations. Academic Press, London
—(1984) Irish: planning and preservation. Journal of Multilingual and Multi¬
cultural Development, 5, Nos 3 & 4, 267—75
—(1985) Language, Society and Identity, Basil Blackwell, Oxford
ELIAS-OLIVARES, L and NASJLETI, D (eds) (1983) Spanish in the United States:
Beyond the Southwest, NCBE, Washington
ELWERT, w T (1973) Das zweisprachige Individuum. Ein Selbstzeugnis. Studien
zu den romanischen Sprachen, Band IV, Franz Steiner Verlag, Wiesbaden
EMRICH, L (1938) Beobachtungen liber Zweisprachigkeit in ihrem Anfangsstu-
dium, Deutsch imAusland 21, 419-24
322 REFERENCES
ERVIN, s (1964) An analysis of the interaction of language, topic and listener,

American Anthropology 66 Part 2, 86-102 (also in J Fishman (ed) Readings
in the Sociology of Language, Mouton, The Hague, 1970, 192-211)
ERVIN, s and OSGOOD, c (1954) Second language learning and bilingualism.
Journal of Abnormal and Social Psychology Supplement, 139^6
ERViN-TRipp, s (1964) Language and TAT content in bilinguals. Journal of Ab¬
normal and Social Psychology 68, 500—7
FABRE, P (1979) Los catalanes, Epidauro, Barcelona
FANTINI, A (1973) Social cues and language choice, in P Turner, (ed) Bilingual¬
ism in the South West, University of Arizona Press, Tucson, 87-111 (revised
edn 1982)
—(1976) Language Acquisition of a Bilingual Child: a sociolinguistic perspec¬
tive, The Experimental Press, Battleboro, Vermont
—(1982) La adquisicion del lenguaje en un niho bilingue. Editorial Herder,
Barcelona
—(1985) The Language Acquisition of a Bilingual Child, Multilingual Matters,
Clevedon
FASOLD, R (1984) The Sociolinguistics of Society, Basil Blackwell, Oxford
FELIX, s (ed) (1980) Second Language Development: Trends and Issues, Gunter
Narr Verlag, Tubingen
FERGUSON, c (1959) Diglossia, Word 15, 325^0
—(1968) Language development, in Fishman, J, Ferguson, C and Das Gupta
(eds) Language Problems of Developing Nations, John Wiley, New York,
27-36
FERRER I GIRONES (1985) La persecucio poUtica de la llengua catalana, Edi-
ciones 62, Barcelona
FISHMAN, J (1964) Language maintenance and language shift as fields of in¬
quiry, Linguistics 9, 32-70
—(L965a) Who speaks what language to whom and when? La Linguistique, 2,
67-88 (revised 1971 version in Pride and Holmes (eds) Sociolinguistics 15-
33)
—(1965b) Bilingualism, intelligence and language learning. Modern Language
Journal 49, 227-37
—(1966a) The implications of bilingualism for language teaching and language
learning, in A Valdman (ed) Trends in Language Teaching, McGraw-Hill,
New York, 146-58
—(1966b) Language Loyalty in the United States, Mouton, The Hague
—(1967) Bilingualism with and without diglossia; diglossia with and without
bilingualism. Journal of Social Issues xxm No 2, 29-37
—(1968) Bilingualism in the Barrio, Pinal Report, US Department of Health,
Education and Welfare, Office of Education, Bureau of Research, vol 2
(August). Also in J Fishman (1971) Advances in the Sociology of Language
1, Mouton, The Hague
—(ed) (1970) Readings in the Sociology of Language, Mouton, The Hague
REFERENCES 323
—(1971) National languages and languages in wider communication in the

developing nations, in W H Whiteley (ed) Language Use and Social
Change, OUP London, 27-56
—(ed) (1972a) Advances in the Sociology of Language, 2 vols, Mouton, The
Hague
—(1972b) Language and Nationalism: Two Integrative Essays, Newbury
House, Rowley, Mass
—(1972c) Language maintenance and language shift, in The Sociology of Lan¬
guage, 76-134
FISHMAN, J, FERGUSON, c and DAS GUPTA (eds) (1968) Language Problems of
Developing Nations, John Wiley, New York
FOSTER, c R (ed) (1980) Nations without a State, Praeger, New York
FRANCIS, E (1947) The nature of the ethnic group, American Journal of Socio¬
logy 52, 393-400
—(1976) Interethnic Relations, Elsevier, New York
FRITSCHE, M (1982) Mehrsprachigkeit in Gastarbeiterfamilien, Deutsch auf der
Basis tiirkischer Syntax, in K Bausch (ed): Mehrsprachigkeit in der Stadt-
region, IdS Jahrbuch 1981, Padagogischer Verlag Schwann, 160-224
—(1984) Remigration/Reintegration - ein Gegenstand der Sprachwissenschaft?
Informationen Deutsch als Fremdsprache, 3, Jahrgang 84/85, August 1984,
DAAD, 5-8
FUNDACIO JAUME BOFILL (1989) Catalunya 77-88, Barcelona
GAL, S (1978a) Peasant men can’t get wives: language change and sex roles in
a bilingual community. Language in Society, 1, No. 1, 1-16
—(1978b) Variation and change in patterns of speaking: language shift in Aus¬
tria, in D Sankoff (ed) Linguistic Variation: Models and Methods, Academic
Press, New York, 227-38
—(1979) Language Shift: social determinants of linguistic change in bilingual
Austria, Academic Press, New York
GANS, H (1979) Symbolic ethnicity: the future of ethnic groups and cultures in
America, Ethnic and Racial Studies 2, 1-20
GARDNER, J w (1965) Foreword to: C E Osgood and T A Sebeok (eds) Psycho¬
linguistics: a survey of theory and research problems, Indiana University
Press, Bloomington
GARDNER, R c and LAMBERT, w E (1972) Attitudes and motivation in second-
language learning, Newbury House, Rowley, Mass.
GARDNER-CHLOROS, P (1983) Hans im Schnockeloch: Language in Alsace.
Modern Languages, 64(1), 35^1
—(1985) Language selection and switching among Strasbourg shoppers. Inter¬
national Journal of the Sociology of Language, 54, 117-35
GARVIN, P and MATHIOT, M (1970) The urbanization of the Guarani language: a
problem in language and culture, in J Fishman (ed) (1970) Readings in the
Sociology of Language, Mouton, 365-74
324 REFERENCES
GAUTHIER, N (1982) Alsace, in R Lafont (ed) Langue dominante, langue domi-

nee, Edilig, Paris, 73-83
GEISELBERGER, s (1972) Schwarzb'uch: Ausldndische Arbeiter, Herausgegeben
im Auftrag des Bundesvorstandes der Jungsozialisten, Fischer, Frankfurt
GENESEE, F (1987) Learning through two languages, Newbury House, Cam¬
bridge, Mass.
—(1989) Farly Bilingual Language Development: One Language or Two?,
Journal of Child Language 16, 1, 161-79
GERTH, H H and MILLS, C w (eds) (1948) Mox Weber, Essays in sociology. Rout-
ledge & Kegan Paul, London
GIBBONS, J (1983) Attitudes towards languages and code-mixing in Hong Kong,
Journal of Multilingual and Multicultural Development 4, Nos 2 & 3, 129-
A1
—(1986) Code-mixing and Code-choice, Multilingual Matters, Clevedon
GILES, H (ed) (1977) Language, Ethnicity and Intergroup Relations, Academic
Press, London
GILES, H, BOURHIS, R and TAYLOR, D (1977) Towards a theory of language in
ethnic group relations, in Giles, H (ed) Language, Ethnicity and Intergroup
Relations, Academic Press, London, 307-^9
GINER, S (1984) The Social Structure of Catalonia, The Anglo-Catalan Society,
Occasional Publications No 1, London
GOLDBERG, MILTON (1941) A qualification of the Marginal Man theory, Ameri¬
can Sociological Review 6, 52-8
GOLDSTEIN, J and BENVENUE, R (eds) (1980) Ethnicity and ethnic relations in
Canada, Butterworth, Toronto
GREENE, D (1981) The Atlantic group: Neo-Celtic and Faroese, in F Haugen, J
McClure and D Thomson (eds) Minority Languages Today, Edinburgh
University Press, Edinburgh 1-9
GREENFIELD, L (1972) Situational measures of normative language views in re¬
lation to person, place and topic among Puerto Rican bilinguals, in J
Fishman (ed) Advances in the Sociology of Language, vol ll, Mouton,
17-35
GROSJEAN, F (1982) Life with Two Languages, Harvard University Press, Cam¬
bridge, MA
—(1985a) The bilingual as a competent but specific speaker-hearer. Journal of
Multilingual and Multicultural Development, 6 No 6, 1985, 467-77
—(1985b) Polyglot aphasics and language mixing: a comment on Perecman
1984, Brain and Language 26, 349-55
GUMPERZ, J (1964) Linguistic and social interaction in two communities,
American Anthropologist 66, No 2, 37-53
—(1977) Social network and language shift, in C Molony, H Zobl and W
Stolting (eds) Deutsch im Kontakt mit anderen Sprachen, German in Con¬
tact with other Languages, Scriptor Verlag Kronberg/Ts
REFERENCES 325
GUMPERZ, j and hernAndez-chAvez, e (1972) Bilingualism, bidialectalism and

classroom interaction, in J J Gumperz Language in Social Groups, Stanford
University Press, 311-39
—(1975) Cognitive aspects of bilingual communication, in E Hernandez-Cha-
vez, et al. (eds) El lenguaje de los chicanos. Center for Applied Linguistics,
Arlington, 154—63
GUMPERZ, J and HYMES, D (eds) (1986) Directions in Sociolinguistics, Basil
Blackwell, Oxford
HALLIDAY, M A K (1975) Learning how to Mean: explorations in the develop¬
ment of language, Edward Arnold, London
HALLIDAY, M A K, MCINTOSH, A and STREVENS, P (1970) The users and uses of
language, in J Fishman (ed) Readings in the Sociology of Language, Mou-
ton. The Hague 137-69
HAMERS, J and BLANC, M (1989) Bilinguality and Bilingualism, Cambridge
University Press (first published in French by Pierre Mardaga, Liege, 1983)
hansegArd, n e (1968) Tvdsprdkighet eller halvsprdkighet? Aldus series,
Stockholm (3rd edn 1972)
HARDING, E and RILEY, p (1986) The Bilingual Family, Cambridge University
Press
HARRISON, G J and PIETTE, A B (1980) Young bilingual children’s language selec¬
tion, Journal for Multilingual and Multicultural Development 1 No 3,
217-30
HARSHMAN, R and KRASHEN, s (1972) An ‘unbiased’ procedure for comparing
degree of lateralization of dicholically presented stimuli. Journal of the
Acoustical Society of America, 52, 174 (Abstract)
HARTMANN, c (1980) Der Sprachkonflikt in Catalunya-Nord aus der Perspek-
tive engagierter Katalanisten, PhD thesis. University of Frankfurt
HATCH, E (ed) (1978) Second Language Acquisition: a book of Readings, New¬
bury House, Rowley
HAUG, E (1984) Alsace und Elsass: Beitrdge zur gegenwdrtigen Identitdtskrise
einer europdischen Landschaft, Pfaehler
HAUGEN, E (1953) The Norwegian Language in America, University of Penn¬
sylvania Press, Philadelphia
—(1956) Bilingualism in the Americas: a bibliography and research guide.
Publications of the American Dialect Society, 26, Alabama
—(1966a) Dialect, language, nation, American Anthropologist 68, 930-40
—(1966b) Language Conflict and Planning: the case of modern Norway, Har¬
vard University Press, Cambridge, MA
—(1966c) Linguistics and language planning, in W Bright (ed) Conference of
Sociolinguistics 1964 Los Angeles, Mouton, The Hague 50-71
—(1969) Language planning, theory and practice, in Actes du Congres In¬
ternational des Linguistes, Bucharest 1967, edited by A. Graur, Editions de
L’Academie de la Republique Socialiste de Roumanie, Bucharest, 701-11
326 REFERENCES
—(1983) The implementation of corpus planning: theory and practice, in J

Cobarrubias and J Fishman (ed.s) Progress in Language Planning, Mouton
de Gruyter, Berlin 1983, 269-90
HAUGEN, E, MCCLURE, D and THOMSON, D (eds) (1981) Minority Languages
Today, Edinburgh University Press, Edinburgh
HEATH, s B and MANDABACH, F (1983) Language Status decision and the law in
the United States, in J Cobarrubias and J Fishman (eds) Progress in Lan¬
guage Planning, Mouton de Gruyter, Berlin, 87-106
HERAUD, G (1968)Peuples et langues d’Europe, Presses d’Europe, Paris
—(1974) L’Europe des ethnies. Presses d’Europe, Paris (first edn 1963)
—(1979-82) Le statut des langues dans les differents etats et en particulier en
Europe, Annales de la Faculte de Droit de TUniversite de Toulon et du Var
5, 81-106
hernAndez-chAvez, e, COHEN, A D and BELTRAMO, A F (eds) (1975) El lenguaje
de los chicanos. Center for Applied Linguistics, Arlington
hernAndez PINA, F (1984) Teorlas psico-sociolingmsticas y su aplicacion a la
adquisicion del espahol como lengua materna, Siglo Veintiuno, Madrid
HEWITT, R (1982) White adolescent Creole users and the politics of friendship.
Journal of Multilingual and Multicultural Development, 3, No 3, 217-32
HINDERLING, R (ed) (1986) Europdische Sprachminderheiten im Vergleich,
Franz Steiner Verlag, Stuttgart
HOFFMANN, c (1979) Bilingualism in a two-year-old child. Polyglot I, No. 2
—(1985) Language Acquisition in two trilingual children. Journal of Multilin¬
gual and Multicultural Development, 6, No 6, 479-95
—(1988) Linguistic normalization in Catalonia; Catalan for the Catalans or
Catalan for Catalonia? in J N Jprgensen et al. (ed) Bilingualism in Society
and School, Copenhagen Studies in Bilingualism, 5: 33-44, Multilingual
Matters, Clevedon
—(1'989) Modelos de Adquisicion del bilingiiismo infantil, Actas del VI Con-
greso Nacional de Lingiiistica Aplicada, Universidad de Santander 17^5
—(forthcoming) Problems encountered in the description of early bilingual
competence
HOMEL, P, PALIJ, M and AARONSON, J (eds) (1987) Childhood Bilingualism, Law¬
rence Erlbaum Associates, New Jersey
HOOPER, J (1986) The Spaniards, Viking (Penguin Books), Harmondsworth
HORNBY, p (ed) (1977) Bilingualism, Academic Press, London
HOUSTON, s H (1972) Bilingualism: naturally acquired bilingualism, in A Survey
of Psycholinguistics, Mouton, The Hague, 203-25
HOYER, A E and HOYER, G (1924) Uber die Lallsprache eines Kindes, Zeitschrift
fiir angewandte Psychologic 24, 363-84
HUSEN, T and OPPER, s (eds) (1983) Multicultural and Multilingual Education in
Immigrant Countries, Wenner Gren Symposium Series vol 38, Pergamon
Press, Oxford
REFERENCES 327
lANCO-WORRALL, A D (1972) Bilingualism and cognitive development, Child

Development 43, 1390-400
IMEDADZE, N V (1967) On the psychological nature of child speech formation
under conditions of exposure to two languages, International Journal of
Psychology, 2, 129-32
ISAJIW, w (1980) Definitions of ethnicity, in J Goldstein and R Bienvenue (eds)
Ethnicity and ethnic relations in Canada, Butterworth, Toronto
ITOH, H and HATCH, E (1978) Second language acquisition: a case study, in E
Hatch (ed) Second Language Acquisition: a book of readings, Newbury
House, Rowley
JAHN, L F (1808) Deutsches VoUcstum (reprinted in Hint’s Deutsche Sammlung,
Breslau 1930)
jAKOBOvrrs, l a (1968) Dimensionality of compound-coordinate bilingualism.
Language Learning 3, 29-56
JAKOBSON, R (1941) Kindersprache, Aphasie und allgemeine Lautgesetze, Alm-
quist und Wiksell, Uppsala
JERNUDD, B (1973) Language planning as a type of language treatment, in J
Rubin and R Shuy (eds) Language Planning: current issues and research,
Georgetown University Press, Washington DC, 11-23
JONES, w R (1953) The influence of reading ability on intelligence test scores of
Welsh-speaking children, British Journal of Educational Psychology 23,
114-20 ^
—(1955) Bilingualism and reading ability in English, University of Wales
Press, Cardiff
—(1959) Bilingualism and intelligence, University of Wales Press, Cardiff
JORGENSEN, J N (1984) Eour kinds of bilingualism in an industrialized society,
Sociolinguistic Papers, Royal Danish School of Educational Studies,
Copenhagen
KEDOURIE, E (1961) Nationalism, Hutchinson, London
KEIM, I (1984) Untersuchungen zum Deutsch tiirkischer Arbeiter, Gunter Narr,
Tubingen
KELLER, H (ed) (1976) Bilingualism in the Centennial and Beyond, Bilingual
Press - Prensa Bilingiie, New York
KELLY, L (ed) (1969) The Description and Measurement of Bilingualism,
University of Toronto Press
KESSLER, c (1971) The Acquisition of Syntax in Bilingual Children, George¬
town University Press, Washington DC
—(1972) Syntactic contrasts in child bilingualism. Language Learning 22,
221-33
—(1984) Language acquisition in bilingual children, in N Miller (ed) Biling¬
ualism and Language Disability, Croom Helm, London, 26-54
KIELHOFER, B and JONEKEIT, S (1983) Zweisprachige Kindererziehung, Staffen-
burg Verlag, Tubingen
328 REFERENCES
KLEIN, w and DITTMAR, N (1979) Developing Grammars. The Acquisition of

German Syntax by Foreign Worlêrs, Springer, Berlin
KLEIN, w (1984) Zweitsprachenverh. Fine Einfuhrung, Athenaum, Konig-
stein/Ts
—(1985) Second Language Acquisition, Cambridge University Press, Cam¬
bridge
KLOSS H (1967a) Types of multilingual communities: a discussion of ten vari¬
ables, in S Liebersohn (ed) Explorations in Sociolinguistics, Mouton, The
Hague, 7-17
—(1967b) Bilingualism and nationalism. Journal of Social Issues xxni No 2,

39^7
—(1968) Notes concerning a language-nation typology, in J Fishman, C Fergu¬
son and Das Gupta J (eds) Language Problems of Developing Nations, John
Wiley, New York, 69-86
—(1984) Umriss eines Forschungprogrammes zum Thema ‘Sprachentod’, In¬
ternational Journal of the Sociology of Language 45, 65—76
KRASHEN, s D (1975) The development of cerebral dominance and language
learning: more new evidence, in D P Dato (ed) Georgetown University
Round Table on Language and Linguistics, Georgetown University Press,
Washington DC
KUTSCH, s and DESGRANGES, I (eds) (1985) Zweitsprache Deutsch - ungesteuer-
ter Erwerb Interaktionsorientierte Analysen des Projekts Gastarbeiter-
kommunikation, Niemeyer, Tubingen
LABOV, w (1963) The social motivation of a sound change. Word 19: 273-309
(also in Labov 1972 Sociolinguistic Patterns)
—(1966) The Social Stratification of English in New York City, Center for
Applied Linguistics, Washington
—(1972) Sociolinguistic Patterns, University of Pennsylvania Press, Philadel¬
phia
LADIN, w (1982) Der elsdssische Dialekt - museumsreif SAUTE (Societe Alsa-

cienne et Lorraine de Diffusion et d’Edition), Strasbourg
—(1983) Mehrsprachigkeit in Ostfrankreich, Sprachverhalten und Sprach-
bewuBtsein der unterelsassischen dialektophonen Schuljugend, in P H Nelde
(ed) Plurilingua iv, Ferdinand Diimmlers Verlag, Bonn
LADO, R (1957) Linguistics across Cultures, University of Michigan, Ann Arbor
LAMBERT, w E (1972) Language, Psychology and Culture, Stanford University
Press, Stanford
—(1974) Culture and language as factors in learning and education, in F

Aboud and R D Mead (eds) Cultural Factors in Learning, Washington
State College, Bellingham
—(1977) The effects of bilingualism on the individual, in P Hornby (ed) Bi¬

lingualism, Academic Press, London 15-27
LAMBERT, w E, havelkA, J and GARDNER, R c (1959) Linguistic manifestations
of bilingualism, American Journal of Psychology 72, 77-82
REFERENCES 329
LAMBERT, w E, JUST, M and SEGALOWITZ, N (1970) Some cognitive consequences

of following the curricula of the early school grades in a foreign language,
in J E Alatis (ed) Georgetown University Monograph Series on Languages
and Linguistics, 23 Georgetown University Press, Washington DC 229-79
LAMBERT, w E and TUCKER, G R (1972) Bilingual education of children: the St
Lambert experiment, Newbury House, Rowley, Mass.
Lectures on Language and Linguistic Method in the School,
LAURIE, s s (1890)
Cambridge University Press, Cambridge
LENNEBERG, E H (1967) Biological Foundations of Language, Wiley, New York
LEOPOLD, w F (1939, 1947, 1949a, 1949b) Speech Development of a Bilingual
Child: a linguist’s record, 1 Vocabulary growth in the first two years; II
Sound learning in the first two years, /// Grammar and general problems in
the first two years, iv Diary from age two. Northwestern University Press,
Evaston
LEPSCHY, A and LEPSCHY, G (1977) The Italian Language Today, Hutchinson
University Library, London
LEWIS, G E (1972) Multilingualism in the Soviet Union, Mouton, The Hague
LIEBERSON, s (1972) Bilingualism in Montreal: a demographic analysis, in J
Fishman (ed) Advances in the Sociology of Language, Mouton, The Hague,
231-54
LINDHOLM, K J and PADILLA, A M (1978a) Child bilingualism: report on language
mixing, switching and translations. Linguistics 16, 23^4
—(1978b) Language mixing in bilingual children. Journal of Child Language
5, No 2, 327-35
LINGUISTIC MINORITIES IN ENGLAND, (1983) a report by the Linguistic Mi¬
norities Project for the Department of Education and Science, Institute of
Education, London. Tinga Tinga, Heinmann
LIPSKI, J (1978) Code-switching and the problem of bilingual competence, in M
Aspects of Bilingualism, Hornbeam, Columbia, 250—60
Paradis (ed)
lUdi, G and PY, BERNARD (1984) Zweisprachig durch Migration: Einfuhrung in
die Erforschung der Mehrsprachigkeit am Beispiel zweier Zuwanderer-
gruppen in Neuenburg, Max Niemeyer Verlag, Tubingen
MACKEY, w F (1962) The description of bilingualism, Journal of the Canadian
Linguistic Association, 1, 51-85

—(1970) The description of bilingualism, in J Fishman (ed) Readings in the
Sociology of Language, Mouton, The Hague 554—84
_(1983) Language status policy and the Canadian experience, in J Cobarru-
bias and J Fishman (eds) Progress in Language Planning, Mouton de
Gmyter, Berlin, 173-206

MACNAB, G L (1979) Cognition and bilingualism: a reanalysis of studies. Lin¬
guistics 17, 231^55
MACNAMARA, J (1967a) Bilingualism in the modern world. Journal of Social
Issues, XXIII, 2, 1—7
330 REFERENCES
—(1967b) The bilingual’s linguistic performance - a psychological overview,

Journal of Social Issues xxill, 2, 58-77
—(1967c) The effects of instruction in a weaker language, Journal of Social
Issues XXIII, 2, 121-35
—(1969) How can one measure the extent of a person’s bilingual proficiency?,
in L. Kelly (1969) Description and Measurement of Bilingualism, Univer¬
sity of Toronto Press, 80-98
—(1971) Successes and failures in the movement for the restoration of Irish, in
J Rubin and B Jemudd (eds) Can Languages be Planned? University Press
of Hawaii, Honolulu
MARSHALL, D (1986) The question of an official language: language rights and
the English Language Amendment, International Journal of the Sociology
of Language 60, 7-71
MARTIN-JONES, M and ROMAINE, s (1986) Semilingualism: a half-baked theory of
communicative competence. Applied Linguistics, 1 1, 26-36
MCCLURE, E (1977) Aspects of code-switching in the discourse of bilingual
Mexican-American children. Technical Report No 44, Center for the Study
of Reading, University of Illinois, Urbana-Champaign
MCLAUGHLIN, B (1984) Second-Language Acquisition in Childhood, vol I Pre¬
school Children (2nd edn), Hillsdale, London (first published by Lawrence
Erlbaum Associates, New Jersey, 1978)
MEISEL, j (1975) Auslanderdeutsch und Deutsch auslandischer Arbeiter. Zur
mbglichen Entstehung eines Pidgin in der BRD. Zeitschrift fur Literatumis-
senschaft und Linguistik, May, 18, 7-53
—(1977) The language of foreign workers in Germany, in Molony, Zobl and
Stolting (eds): Deutsch im Kontakt mit anderen Sprachen, German in Con¬
tact with other Languages, Kronberg/Ts, Scriptor Verlag, 184-212
—(1984) Zum Erwerb von Wortstellungsregularitaten und Kasusmarkiemngen,
Monograph, Universitat Hamburg Romanisches Seminar
—(1986) Word order and case marking in early child language: evidence from
simultaneous acquisition of two first languages (Erench and German), Lin¬
guistics 24, 123-83
—(1987) Early differentiation of languages in bilingual children, in K Hylten-

stam and L Obler (eds): Bilingualism across the Lifespan: in health and in
pathology, Cambridge University Press, Cambridge, 13^0
METRAUX, R w (1965) Study of bilingualism among children of US-French
parents, French Review 38, 650-65
MIKES M (1967) Acquisition des categories grammaticales dans le langage de

1’enfant, Enfance 20, 289-98
MILLER, J (1983) Many Voices, Routledge & Kegan Paul, London
MILLER, N (ed) (1984) Bilingualism and Language Disability, Croom Helm,
London
MIRACLE, A (ed) (19S3) Bilingualism: social issues and policy implications.

University of Georgia Press
REFERENCES 331
MKILIFI, M (1978) Triglossia and Swahili-English bilingualism in Tanzania, in J

Fishman (ed) Advances in the Study of Societal Multilingualism, Mouton,
The Hague
MOLONY c, ZOBL H and STOLTING w (eds) (1977) Deutsch im Kontakt mit
anderen Sprachen, German in Contact with other Languages, Scriptor
Verlag, Kronberg/Ts
MURRELL, M (1966) Language acquisition in a trilingual environment, Studia
Linguistica 20, 9-35
NELDE, p (1984) Three issues in languages in contact, Studia Anglica
Posnaniensia 17, 147-56
NELDE, P et al. (eds) (1981) Sprachprobleme bei Gastarbeiterkindern, Gunter
Narr, Tubingen
Monographs of the
NELSON, K (1973) Structural and strategy in learning to talk,
Society for Research in Child Development 39 (1-2 serial No 149)
NEUSTUPNY J (1970) Basic types of treatment of language problems. Linguistic
Communications 1, 77—98
NEVILLE, G (1986) Minority languages in contemporary France, Journal of
Multilingual and Multicultural Development, 8, Nos 1 & 2, 147-57
NIKOLINAKOS, M (1973) Politische Okonomie der Gastarbeiterfrage. Migration
und Kapitalismus, Reinbeck, Rowohlt Verlag
OESTREICHER, J P (1974) The early teaching of modem language, education and
culture. Review of the Council for Cultural Cooperation of the Council of
Europe 24, 9—16
OKSAAR, F (1970) Zum Spracherwerb des Kindes in zweisprachiger Umge-
bung. Folia Linguistica 4, 330—58
—(1971) Spr^kpolitiken och minoritetema, in D Schwarz (ed) Identitet och
minoritet, Teori och praktik i dagens Sverige, Almquist och Wiksell, Stock¬
holm, 164—75
—(1974) On code-switching: an analysis of bilingual norms, in J Quistgaard,
H Schwarz and Spang-Hansen, H (eds) Applied Linguistics: problems and
solutions. Proceedings of the Third AILA Congress, Copenhagen, Heidel¬
berg, Vol 3, 491-500
—(1980) The multilingual language acquisition project. International Review
of Applied Psychology 29, 268-9
—(1983) Multilingualism and multiculturalism from the linguist’s point of
view, in T Husen and S Opper (eds) Multicultural and Multilingual Educa¬
tion in Immigrant Countries 17-36
—(ed) (1987) Soziokulturelle Perspektiven von Mehrsprachigkeit und
Spracherwerb - Sociocultural Perspectives of Multilingualism and Lan¬
guage Acquisition, Gunter Narr, Tubingen
OOMEN-WELKE, I (1985) Eifahrungen und Reflexionen aus der Arbeit in einem
‘nationalen Model!’, Deutsch Lernen Heft 2
332 REFERENCES
ORLOVIC-SCHARWARZWALD, M (1978) Zum Gastarbeiterdeutsch jugoslawischer

Arbeiter im Rhein-Main Gebiet. Mainzer Studien zur Sprach- und Volks-
forschung, 2, Wiesbaden
PADILLA, A M and LIEBMAN, E (1975) Language acquisition in the bilingual child,
Bilingual Review 2, 34-55
PARADIS, M (1977a) Bilingualism and aphasia, in H A Whitaker and H Whitaker
(eds) Studies in Neurolinguistics, 3, Academic Press, New York, 65-121
—(1977b) The stratification of bilingualism, in R J di Pietro, and E L Balnsitt
(eds) The Third Lacus Forum 1976, Hornbeam, Columbia, 231-41
—(ed) (1978) Aspects of Bilingualism, Hornbeam, Columbia
PARADIS, M and LEBRUN, Y (eds) (1984) Early Bilingualism and Child Develop¬
ment, Swets & Zeitlinger, Lisse
PARASHER, s N (1980) Mother-tongue English diglossia; a case study of edu¬
cated Indian bilinguals’ language use. Anthropological Linguistics 22, 4,
151-68
PAULSTON, c B (1974) Questions concerning bilingual education, paper
presented at the Interamerican Conference on Bilingual Education, 1974
PAVLOVITCH, M (1920) Le langage enfantin: acquisition du serbe et du frangais
par un enfant serbe. Champion, Paris
PEAL, E and LAMBERT, w E (1962) The relationship of bilingualism to intel¬
ligence, Psychological Monographs 76 (27), 1-23
PENFIELD, w and ROBERTS, L (1959) Speech and Brain Mechanisms, Princeton
University Press, Princeton
PENN, D (1990) Minority languages in Europe, The Linguist, Vol 29 No 2, 62-9
PETERS, A M (1977) Language-learning strategies: does the whole equal the sum
of the parts? Language 53, 560-73
c (1979) Constraints on language mixing: intersentential code-switching
PFAFF,
and borrowing in Spanish/English, Language 55, 291-318 (also in J Amas-
tae and L Elfas-Olivares (eds) (1982) Spanish in the United States:
sociolinguistic aspects, Cambridge University Press, Cambridge, 264-97)
—(1980) Acquisition and development of ‘Gastarbeiterdeutsch’ by migrant
workers and their children in Germany, in E Trangott et al. (eds): Papers
from the Conference on Historial Linguistics, Amsterdam Studies in the
Theory and History of Linguistic Sciences, iv, vol 14. John Benjamin BV,
Amsterdam, 381-95
—(1981) Sociolinguistic problems of immigrants: foreign workers and then-
children in Germany, Language in Society, 10, 155-88
PFAFF, c and PORTZ, R (1980) Children’s acquisition of German: universal vs.
interference, in Proceedings of the Second Scandinavian-German Sympo¬
sium of the language of migrant workers and their children. West Berlin.
(Also: Linguistische Arbeiten und Berichte, Heft 16, Berlin Ereie Universi-
tat, 137-70)
PHILIPPS, E(1975) Les luttes linguistique en Alsace jusqu’en 1945. Culture Al-
sacienne, Grisheim-sur-Souffel
REFERENCES 333
—(1978) La arise d’identite (UAlsace face a son destin). Societe d’edition de

la Basse Alsace, Strasbourg
PICHON, E (1936) Le Developpement Psychique de I’Enfant et de 1’Adolescent,
Masson, Paris
PIENEMANN, M (1980) The second language acquisition of immigrant children,
in S Felix (ed): Second Language Development, Gunter Narr, Tubingen
POKE, J (1965) Bilinguismes, Revue Roumaine de Linguistique 10, 343-9
POPLACK, s (1980) Sometimes I’ll start a sentence in English y termino en
espanol: towards a typology of code-switching. Linguistics 18, 582-618
(also in J Amastae and L Elias-Olivares (eds) (1982) Spanish in the United
States, Cambridge University Press, Cambridge, 230-63)
PORSCHE, D (1983) Die Zweisprachigkeit wdhrend des primdren Spracherwerbs
Tiibinger Beitrage zur Linguistik 218, Gunter Narr, Tubingen
POWER, J (1984) Western Europe’s Migrant Workers (in collaboration with A
Hardman; revised by C Byrne-Vardin and K Stearman). Minority Rights
Group Report No 28, Pergamon Press, London
RAFELER-ENGEL, w (1965) Del bilinguismo infantile.. Arch. Glottologico
Italiano 50, 175-80
RATH, R (ed) (1983) Sprach- und Verstdndnisschwierigkeit bei Ausldnderkin-
dern in Deutschland. Aufgaben und Probleme einer interaktionsorientierten
Zweitspracherwerbsforschung, M Lang,
REDLINGER, w G and PARK T (1980) Language mixing in young bilinguals. Jour¬
nal of Child Language 7, 337-52
REHBEIN, J (1985) Interkulturelle Kommunikation, Gunter Narr, Tubingen.
—(1987) Sprachloyalitat in der Bundesrepublik. Auslandische Kinder zwischen
Sprachverlust und zweisprachiger Erziehung. Arbeiten zur Mehrsprachigkeit
26, Universitat Hamburg
REYES, R (1982) Language mixing in Chicano Spanish, in J Amastae and G
Elias-Olivares (eds) Spanish in the United States: sociolinguistic aspects,
Cambridge University Press, Cambridge, 154-65
RITCHIE, w c (1978) Second Language Acquisition Research, Academic Press,
New York
RIS, R(1979) Dialekte und Einheitssprache in der deutschen Schweiz, Interna-
tionalJournal of the Sociology of Language 21, 41-61
ROBERTS, M (1939) The problem of the hybrid language. Journal of English
and Germanic Philology 38, 23-41
ROMAINE, S (1984) The Language of Children and Adolescents, Basil Black-
well, Oxford
—(1989) Bilingualism, Basil Blackwell, Oxford
RONJAT, J (1913) Le developpement du langage observe chez un enfant bi-
lingue. Champion, Paris

(1968b) National Bilingualism in Paraguay, Mouton, The Hague
RUBIN, J
—(1970) (ed.) Bilingual usage in Paraguay, in Fishman (1970). Readings in the
Sociology of Language, Mouton, The Hague, 512—30
334 REFERENCES
—(1973) Language planning: discussion of some current issues, in J Rubin

and R Shuy (eds) Language Planning: Current Issues and Research,
Georgetown University Press, Washington D C, 1-10
rUKE-DRAVINA, V (1965) The process of acquisition of apical /r/ and uvular /r/
in the speech of children. Linguistics 17, 56—68
—(1967) Mehrsprachigkeit im Vorschulalter, Traveaux de LTnstitut de Phone-
tique de Lund, Ease, v, Lund Gleerup
RUSS, c and VOLKMAR, c (eds) (1987) Sprache und Gesellschaft in deutsch-
sprachigen Ldndern, Goethe Institut, Miinchen
SABATER, E (1984) An approach to the situation of the Catalan language: social
and educational use. International Journal of the Sociology of Language,
47, 29-41
SAER, D J (1922) An inquiry into the effects of bilingualism upon the intel¬
ligence of young children. Journal of Experimental Pedagogy 6, 232-40,
266-74
—(1923) The effects of bilingualism on intelligence, British Journal of Psy¬
chology 14, 25-38
SANDER, F (1930) Structures, Totality of Experience and Gestalt, in C Murchi¬
son, (ed) Psychologies of the 1930s Worcester Clark University Press
—(1934) Seelische Stmktur und Sprache, Stmkturpsychologisches zum Zweit-
sprachenproblem, Neue Psychologische Studien 12, 59
SANKOFF, D (ed) (1978) Linguistic Variation: models and methods. Academic
Press, New York
SANKOFF, G (1980) Language use in multilingual societies: some alternate ap¬
proaches, in G Sankoff, The Social Life of a Language, University of
Pennsylvania Press, 26-46 (also in Pride and Holmes Sociolinguistics, Pen¬
guin, Harmondsworth, 1972, 33-51)
—(1980) The Social Life of a Language, University of Pennsylvania Press,
Pljiladelphia
SAUNDERS, G (1982a) Bilingual Children: guidance for the family. Multilingual
Matters, Clevedon
—(1982b) Infant bilingualism: a look at some doubts and objections. Journal
for Multilingual and Multicultural Development 3, No 4, 277-92
—(1988) Bilingual Children: from birth to teens. Multilingual Matters, Cleve¬
don
SCHLAPFER, R (1982) Die Viersprachige Schweiz, Benziger Verlag
SCHLYTER, s (1988) Language mixing and linguistic level in three bilingual
children, Scandinavian Working Papers on Bilingualism No I
SCHMIDT-ROHR, G (1932) Die Sprache als Bildnerin der Volker, Jena Mutter-
sprache, Miinchen
—(1936) Zur Frage der Zweitsprachigkeit, Deutsche Arbeit 36, 408-11
SCHUMANN, J H (1976) Social distance as a factor in second language acquisi¬
tion, Language Learning, 26 (1), 135^3
REFERENCES 335
—(1978) The Pidginization Process: a model for second language acquisition.

Newbuiy House, Rowley, Mass
SCOTT, s (1973) The relation of divergent thinking to bilingualism - cause or
effect? (unpublished research report, McGill University)
SELIGER, H w (1978) Implication of a multiple critical-period hypothesis for
second language learning, in W C Ritchie (ed) Second Language Acquisi¬
tion Research, Academic Press, New York
SHAPSON, s and D’OYLEY V (eds) (1984) Bilingual and Multilingual Education:
Canadian perspectives. Multilingual Matters, Clevedon
SHIBUTANI, T and KWAN, K (1965) Ethnic Stratification, Macmillan, New York
SIGUAN, M (1980) Education and bilingualism in Catalonia, Journal of Multilin¬
gual and Multicultural Development, 1, No 3, 231^2
—(1989) Escribir y pensar en dos lenguas. Revista critica de libros Saber Leer,
Eundacidn Juan March, 22 (Eebmary)
silva-CORVAlAn, c (1982) Subject variation in spoken Mexican-American
Spanish, in J Amastae and L Ehas-Olivares, (eds) Spanish in the United
States Cambridge University Press, Cambridge 93-120
—(1983) Code-shifting patterns in Chicano Spanish, in L Ehas-Olivares and D
Nasjleti (eds) Spanish in the United States: Beyond the Southwest, NCBE,
Washington, 69-87
—(1986) Bilingualism and language change: the extension of estar in Los
Angeles Spanish, Language 62, 587-608
—(1989) Sociolinguistica: teoria y andlisis, Alhambra, Madrid
SKUTNABB-KANGAS, T (1976) Bilingualism, semilingualism and school achieve¬
ment, Linguistische Berichte, 45 October
—(ed) (1977) Papers from the Eirst Scandinavian Conference on Bilingualism,
Department of Nordic Languages, University of Helsinki, Series B No 2
—(1984a) Bilingualism or Not, Multilingual Matters, Clevedon (first edn in
Swedish, 1981)
—(1984b) Children of guest workers and immigrants: linguistic and educa¬
tional issues, in J Edwards (ed) Linguistic Minorities, Policies and
Pluralism, Academic Press, London, 17-47
SKUTNABB-KANGAS, T and CUMMINS, J (1988) Minority Education: from shame
to struggle. Multilingual Matters, Clevedon
SKUTNABB-KANGAS, T and TOUKOMAA, P (1976) Teaching migrant childrens
mother tongue and learning the language of the host country in the context'
of the socio-cultural situation of the migrant family. A Report prepared for
UNESCO, Department of Sociology and Social Psychology, University of
Tampere, Research Report 15, Tampere
—(1979) Semilingualism and middle-class bias. Working Papers in Bilingual¬
ism, 19, November 1979, Ontario Institute for studies in Education, 181-96
SMITH, A (1982) Nationalism, ethnic separation and the intelligentsia, in
C Williams (ed) National Separatism, University of Wales Press, Cardiff
336 REFERENCES
SMITH, M E (1935) A Study of the speech of eight bilingual children of the same
family, Child Development 6, 19-25
SOFFIETI, J (1955) Bilingualism and-biculturalism, Journal of Educational Psy¬
chology 46, 222-7
S0NDERgArd, b (1981) Decline and fall of an individual bilingualism. Journal
of Multilingual and Multicultural Development 2, No 4, 297-302
SPOLSKY, B(1986) Language and Education in Multilingual Settings, Multilin¬
gual Matters, Clevedon
STAVANS, A (1990) Code-switching in children acquiring English, Spanish and
Hebrew: A case study, unpublished PhD thesis. University of Pittsburgh
STEINBERG, S (1981) The Ethnic Myth, Atheneum, New York
STEPHENS, M (1976) Linguistic Minorities in Western Europe, Comer Press,
Llandysul
STOLTING, w (1975) Wie die Auslander sprechen: eine jugoslawische Familie,
Zeitschrift fiir Literaturwissenschaft und Linguistik, 38, May 1975: 54-67
—(1980) Die Zweisprachigkeit jugoslawischer Schuler in der Bundesrepublik
Deutschland, Harrassowitz, Wiesbaden
STRUBELL I TRUETA, M (1984a) Llengua i poblacio a Catalunya, Ediciones de la
Magrana, Barcelona
—(1984b) Language and identity in Catalonia, International Journal of the
Sociology of Language, 47, 91-104
—(1985a) The Catalan experience, paper presented at the Conference on Pub¬
lishing in Minority Languages, Aberystwyth, July/August
—(1985b) Social psychological aspects of the language planning process, un¬
published paper
—(1988) L’avui i el dema de la llengua catalana, in: Panordmica de la llengua
catalana, Institut d’Estudis Autonomies, Generalitat de Catalunya,
Quaderns de Treball 19: 19^2
SUBI^TS, M (1980) La utilitzacio del catala entre la precarietat i la normalitza-
cio. Revista critica de libros Saber Leer, Fundacion Juan March, 1
(February)
SWAIN, M (1972) Bihngualism as a first language, (unpublished PhD disserta¬
tion, University of California, Irvine)
SWAIN, M and LAPKIN, E (1982) Evaluating Bilingual Education: a Canadian
case study. Multilingual Matters, Clevedon
SWAIN, M and WESCHE, M (1975) Linguistic interaction: case study of a bilingual
child. Languages Sciences 37, 17-22
TABOURET-KELLER, A (1962) L’acquisition du langage parle chez un petit enfant
en milieu bilingue, Problemes de Psycholinguistique, 8, 205-19
—(1981) Regional languages in France, International Journal of the Sociology
of Language, 29
TAESCHNER, T (1983) The Sun is Eeminine: a study on language acquisition in
childhood. Springer Verlag, Berlin/Heidelberg
REFERENCES 337
TIMM, L (1975) Spanish-English code-switching: el porque and how-not-to,

Romance Philology 28, 473—82
—(1980) Bilingualism, diglossia and language shift in Brittany, International
Journal of the Sociology of Language, 25, 29^1
TOSI, A (1983) Immigration and Bilingual Education, Pergamon Press
TOVAR, A (1982) El bilingiiismo en Espana, in V Demonte and S Ortega (eds)
El bilinguisimo: problemdtica y realidad, Revista de Occidente 10—11,
Extraordinario B, Madrid, 13-22
TRAVIS, L E, JOHNSON w and SHOVER, J (1937) The relation of bilingualism to
stuttering. Journal of Speech Disorders 2, 185-9
TRUDGILL, p (1972) Sex, covert prestige and linguistic change in the urban
British English of Norwich, Language in Society, 1, 179-95
—(ed) (1978) Sociolinguistic Patterns in British English, Edward Arnold,
London
—(ed) (1984) Language in the British Isles, Cambridge University Press,
Cambridge
TURNER, P (ed) (1973) Bilingualism in the Southwest, University of Arizona
Press, Tucson
VALDES FALLIS, G (1976) Social interaction and code-switching patterns: a case
study of Spanish/English alternation, in G D Keller et al. (eds) Bilingualism
in the Bicentennial and Beyond, Bilingual Press/Editorial Bilingile, New
York, 53-85 (also in J Amastae and L Elias-Olivares (eds) Spanish in the
United States, Cambridge University Press, 1982, 209-29)
VALLVERDU, F (1973) El fet linguistic com afet social, Ediciones 62, Barcelona
VERDOODT, A (1972) The differential impact of immigrant Erench speakers on
indigenous German speakers: a case study in the light of two theories, in J
Eishman Advances in the Sociology of Language, vol 2, Mouton, The
Hague
VIHMAN, M M (1982) The acquisition of morphology by a bilingual child: the
whole word approach. Applied Psycholinguistics 3, 141-60
—(1985) Language differentiation by the bilingual child. Journal of Child Lan¬
guage 12 (2), 297-324
VIHMAN, M M and MCLAUGHLIN, B (1982) Bilingualism and second language
acquisition in preschool children, in C J Brainerd and M Pressley (eds)
Progress in Cognitive Development Reasearch: verbal processes in children
Springer, Berlin, 36-58
VILA, I (1984) La aparicion y el desarrollo de terminos equivalentes en el
primer lenguaje del nino bilingiie familiar, Actas del II Congreso Nacional
de Linguistica Aplicada, Universidad de Murcia, 461-76
VOLTERRA, V and TAESCHNER, T (1978) The acquisition of language by bilingual
children. Journal of Child Language 5, 2, 311-26
VYGOTSKY, L S (1962) Thought and Language, MIT Press, Cambridge, Mass
338 REFERENCES
WALKER, A G H(1987) Language and society in Schleswig, in C Russ and C

Volkmar (eds) Sprache und Gesellschaft in deutschsprachingen Ldndern,
Yorker Werkheft, Goethe Institut, Miinchen, 136-52
WANDRUSZKA, M (1979) Die Mehrsprachigkeit des Menschen, Pieper Verlag,
Miinchen
WARDHAUGH, R (1986) An Introduction to Sociolinguistics, Basil Blackwell,
Oxford
—(1987) Languages in Competition, Basil Blackwell, Oxford
WEBER, M (1968) Economy and Society, Bedminster, New York
WEINREICH, u (1968) Languages in Contact, Mouton, The Hague (first publish¬
ed in 1953)
WEISGERBER, L (1929) Muttersprache und Geistesbildung, Vandenhoek &
Ruprecht, Gottingen
(1933) Die Stellung der Sprache im Aufbau der Gesamtkultur, Worter und
Sachen 15, 134-224; 16, 97-236
WILLIAMS, c H (ed) (1982) National Separatism, University of Wales Press,
Cardiff
(1984) More than the tongue can tell: linguistic factors in ethnic separatism,
in J Edwards (ed) Linguistic Minorities, Policies and Pluralism, Basil
Blackwell, Oxford, 179-219
WOLCK, w (1984) Komplementiemng und Fusion, Prozesse natiirlicher Zwei-
sprachigkeit, in E Oksaar (ed) Spracherwerb-Sprachkontalct-Sprachkonflikt,
De Gruyter, Berlin, 105-29
WOLFF, J(1984) Fremd im eigenen Land. Deutsch als Fremdsprache und
Remigration, Informationen Deutsch als Fremdsprache 3, Jahrgang 84/85
DAAD, 65-73
WOOLARD, K A (1989) Double Talk: Bilingualism and the politics of ethnicity
in Catalonia, Stanford University Press, Stanford
Subject Index
accent, 37, 53, 58 -acquisition (development,

foreign -, 94, 96-7 achievement), 21,44—6,
Act to Promote the French Language 48-52, 60-5, 69-72, 75-9,
in Quebec, 211 106-9, 130, 304-6
age (in bilingualism), 18, 36-8 -adults, 18, 94-5, 96-7, 102,
ambilingualism (equilingualism), 103^, 113
21-2, 29, 35 - bilingual children: see children
attitudes (attitudinal factors), 25-6, - bilingual or multilingual
30-1, 119-20, 134, 257-9, 270, communities (societies,
281,287-8,302-6 ' groups), 4—6, 43^, 116,
- towards bilingualism, 3-4, 175-7, 179-80, 182-5, 189,
27-8, 244 191-2,
- towards code-switching, 113 -competence, 21,74—93, 131,
- towards languages, 201, 133
215-16, 264-5 -deficit, 121-3, 129, 133
- towards the L2, 133 elite bilinguals, 46
(language) awareness, 68,69,71, - environment, 38-40
79-86, 125 - families, 44-6, 47, 175-6, 276
-individual, 24—5, 175-6
“Balance Theoi^”, 129 natural-, 18-19
BICS (Basic Interpersonal - patterns, 24—5, 27, 40-6,
Communication Skills), 127-8, 149-51
132-3 - profile, 31-2
biculturalism (multiculturalism), - speaker, 13-14, 74—5, 96-7,
28-31, 123,312 178-80
Bilingual Education Act, 1968 - speech, 75-6
(USA), 210, 217 - speech mode, 153
bilingual, - upbringing (see also: bilingual
-ability, 21-7,28,31 acquisition), 136
340 SUBJECT INDEX
bilingualism Charter of the French Language,

achieved 19 1977 (Canada), 211
active-, 24, 150,273 child(ren),
additive-, 21, 119, 122 language (acquisition), 53, 55-6
adult, 18 bilingual, 8, 33-54, 55-64, 75-9,
ascendant -, 21 85-93,94-117, 118-135
ascribed -, 19, 33 majority-, 45, 121, 131
balanced-, 22, 123, 129, 130 (second-generation) migrant -,
child- 18, 33-54, 124 41-2, 118-19, 131, 307-10
compound —, 20 minority -, 8, 14-15, 28, 47-8,
coordinate -, 20 127-31, 133-4
definitions of-, 15-16,24 cognates (=cognate forms), 65,
description of -, 14—15, 18-27 99-100
early- 18,33-6,75-9 cognitive development, 49-50, 71,
functional -, 23-5, 39 108-9, 119, 120, 124, 128, 133, 134
incipient -, 21 cognitive functioning, 120-35
individual-, 13-32 cognitive processes, 60-1,67, 120,
infant-, 18 127-8
late-, 18,33-6 communicative competence
maintenance (or loss) of -, 26, (behaviour), 75, 94
31, 44, 45, 150-1, 175-92 communicative needs (situations),
measuring (= assessing) 65-6, 256-8
bilingualism, 151-3 consequences of bilingualism: see
natural-, 18,33 ejfects of bilingualism
passive -, 24, 150, 273 contact (with language): see
primary -, 18, 33 exposure (to language)
productive - 24 context (of bilingualism) 18-19, 78,
receptive -, 24 83,86-8
recessive-, 21, 150-1 fused contexts (of bilingual
school -, 19 acquisition), 18-19, 86
secondary-, 19,33 “corpus planning”, 205-7
societal-, 13-14 Critical Period Hypothesis, 36-7, 58
subordinative -, 20 CUP (Common Underlying
subtractive-, 21,28, 119, 131 Proficiency), 129
(linguistic) blends, 59 linguistic deficit: see bilingual deficit.
borrowing, 31, 95, 101-3, 115, 117,
206 Developmental Interdependence
Bureau for Unrepresented Nations, Hypothesis, 128, 131
198, 233 dialect (dialect groups), 248-50,
254, 255-9, 260-4, 269, 292
CALP (Cognitive Academic regional -, 253, 254, 292
Language Proficiency), 128,132, differentiation (of languages) - see
133 separation (of languages) 128,
catalanism, 268 132,133
SUBJECT INDEX 341
diglossia, diglossic, 166-9, 175, grammatical level (of analysis), 97-8

180, 254-5, 267, 273
domain analysis, 177-80 holophrastic stage, 55, 61
HPD Project {Heidelberger
educational provision / attainment, Forschungsprojekt Pidgin
7-8, 28, 42-4, 118-23, 130-5, Deutsch), 269
191-2, 217, 273-4, 280-4, 287,
304, 307-15 identification, 26, 27
(bilingual) education (programmes), identity, 186,215,228,231,237-8,
10, 42-4, 46-7, 118-19, 128, 244, 252, 262, 296, 300
131, 135, 209, 217, 243-4, 283, group-, 116,238
309, 311 national (ethnic) -, 193—4,
effects (of bilingualism), 118-19, 199-202, 226-31, 252
121-6, 127, 131-5 immersion programmes (courses),
additive- 119, 123-6 42, 47, 118-19, 131, 283-4
detrimental (negative) -, 118-19, immigrants, 28-30, 40-1, 109-10,
121-2, 131 116, 118, 127, 135, 195, 215,
neutral-, 119,122-3 235, 242, 246-7, 264, 271-2,
emigration, 233^, 291 281, 284, 249-69
environment (context): see linguistic immigration, 40-1, 189-90, 234,
environment 242-4, 271-2, 249-69
- (of learning or acquisition), 60, Institute d’Estudis Catalans, 268
74, 80, 86, 91 Institut fiir Deutsche Sprache, 269
equilingualism: see ambilingualism (bilingualism and) intelligence,
(acquisition of) equivalents, 63-5, 120-6
79, 83, 107 interference, 19,31,36,75,95-101,
ethnic origin (= background, group, 102, 129
community), 109-10, 113, 120, International Schools, 43
194-6, 198, 215, 224-5, 237-8
ethnicity, ethnic identity (see also LI (first language, mother tongue,
identity), 194-6,202,215, native language), 25, 27, 29, 36,
237-8, 244 118, 119, 128-9, 131, 230, 246,
exposure (to language; see also 250, 304, 307-12, 314-15
contact with language), 46-8, 57, L2 (second language, foreign
108, 305-6 language), 25, 29, 34, 36, 46,
118,119, 127-9, 131, 133,
Federal Union of European 302-6, 312
Nationalities (FUEN), 231 LAD (Language Acquisition
(language) functions(s), 23—4, 31, Device), 36-7
79, 98, 202, 209, 240-1 language
- acquisition (development), 31,
Gastarbeiter (see migrant workers) 33-7, 50-64, 108, 302-6
grammar (see also: morphology and - allocation, 205, 207-10, 217
syntax), 66-72, 125, 127 base-, 115
342 SUBJECT INDEX
- behaviour: see linguisitic second -: see L2

behaviour -shift, 6, 21, 89, 176, 186-92,
-choice, 75-93, 124-5, 175-85,. 267-8
272-3 - skills, 37-8, 88, 127-30, 285
- competence, 21, 28-31, 74, 92, standard -, 206, 208, 244, 301
127-35, 255, 301-6 - status, 207-18
- conflict, 4-5, 239 - system, 57, 79, 105, 128, 130,
- contact, 5-6 186
- differentiation: see - use, 254-8, 267-8
differentiation Languages Act 1969 (Canada), 211
- dominance (dominant Law of Linguistic Normalization (in
language), 22, 30, 60, 66, 92, Catalonia), 212, 270, 273, 276-84
115, 215, 254, 264, 301 lexis (lexicon, lexical items, lexical
first -: see LI systems; lexical development),
foreign -: see L2 60-5, 76, 77, 99-100, 101-2,
- function(s), 23^, 26-7, 29, 58, 104, 111, 206
67, 79, 98, 202, 205, 240-1, linguistic
245, 255 - accommodation (convergence,
- group, 181, 189 adjustment), 180-2,274
- input, 55-6, 108, 129 - assimilation, 41, 215, 272
- learning, 35-8, 55-64 - behaviour, 1, 9, 71, 94-5, 109,
- loss: see maintenance and loss 113, 116-17, 176-7, 179- 80,
of bilingualism 254-7
- maintenance: see maintenance -community, 26, 175, 182,
and loss of bilingualism 188-92
minority-, 3^, 43-5, 110, - competence: see language
118-19, 133-5, 182, 185, competence
187-92, 203, 209, 214-15, - convergence: see linguistic
^217, 227, 231-3,239-^1, accommodation
'263^, 274 - deficit: see bilingual deficit
national -, 136, 188, 202, 204-5, -environment, 108,119
208, 213, 226-7, 239-41 - features, 31, 90
official -, 188, 201, 207-14, 217, - interaction, 77, 86
226-7, 239-41, 251, 254 -majorities, 44-6, 116
preferred -, 22, 257-8 -markers, 66, 194
- proficiency: see linguistic - minorities, 5, 7-8, 26-7, 34,
proficiency 40-4, 47-8, 118-19, 133-4,
- planning, 191-2, 204-18, 264, 175-6, 182, 187-8, 191, 199,
llZ-l, 287-8 208-9, 215, 217, 219-47, 255,
- recovery (or restoration), 276, 263, 286-8, 289-313
280, 283-^, 287 - normalization, 273-84
regional -, 202, 208, 226-7, - proficiency, 23-4, 31, 94-5,
239-41, 259 113, 119, 129, 176, 274-6, 280
- rights, 214-15, 244-7 - repertoire, 58-9, 93
SUBJECT INDEX
343
metalinguistic (ability or awareness), (linguistic or cultural) pluralism,

124-5 216, 235-6
migrants, migrant workers (or plurilingualism: see multilingualism
groups, Gastarheiter, 34, 41-2, pragmatic competence, 77
116, 118-19, 135, 159-60, 161, prerequisites (for bilingualism),
164, 170-1, 174, 189-90, 227, 46-8, 79
231, 242-7, 289-315 pressure (to become bilingual), 46-8
migration — see immigration (degree or standards of) proficiency
(code-)mixing, 23, 59, 64, 75-6, 84, — see linguistic proficiency
94-101, 103-9, 113, 115 (bilingual) programmes: see
morphology (morphological bilingual education
development, morphological pronunciation, 56-8, 101, 127
features), 56, 57, 66-9, 74, 78, - features (stress, rhythm,
93, 99, 101, 105-6, 110, 133, intonation), 36, 56-8, 96-7,
205, 265 112, 181, 187
mother tongue: see LI
mother tongue teaching, 40-2, 119, remigration, 297-9
214, 307-15
motivation, 133-4, 272, 287-8, semantics (semantic development;
304-6 meaning), 60-2, 99-100, 124-5
multiculturalism: see biculturalism semilingualism, 27-8, 127-8, 130,
multilingualism, 2, 6-10, 26, 133
157-74, 175-7, 201, 219-24, separation (of languages) (see also:
239-41, 245, 313 differentiation), 56-7, 60, 79-88,
105, 107-8
nation (nation-state), 197-8, 225, Separate Development Hypothesis,
228, 262-3 76-9
nationalism (nationality, national separatism, ethnic or cultural
allegiance), 191, 198-202,225, separateness, 237-9
227-9, 237-8, 252-3, 259, 286-7 sequence of acquisition - see stages
native speaker, 26-7, 53, 97, 227 Servei del Catala, 274
Single Switch Theory, 85
Official Language Act, 1974 Spanish Constitution, 1978, 211-12,
(Canada), 211 270, 273
one-parent-one-language strategy, speech behaviour: see linguistic
44-5, 84, 86,108 behaviour
order and consequence (of bilingual speech community - see linguistic
acquisition), 21, 98 community
spelling, 100-101,206,218
phonology (phonological stages (sequence of language
development, phonological acquisition), 69-72, 82, 107, 117
system), 56-60, 74, 95-7, 101, status planning, 207-10, 212-14
105, 115 (learning) strategies, 58-9, 125
pidgin, 208, 304, 313 SUP (Separate Underlying
344 SUBJECT INDEX
Proficiency), 129 - of code-switches, 111-13

(code-)switching (switches), 23, 31, - of interference, 95-6
75, 84-5, 89, 94-101, 109-17,
257, 273 Unitary Language System
syntax (syntactic development, Hypothesis, 75-9
syntactic features), 56, 62, 67,
69-72, 74, 77-8, 97-8, 105, 205, vernacular (vemacularization), 209,
265,306 212, 213, 216, 265, 269
vocabulary: see lexis
Think Tank Model, 129-30 Voting Rights Act, 1965 (USA), 210
Thresholds Theory, 130-1
transfer(ence), 95,97-8, 123 Welsh Language Society, 214
(loan) translation, 94 Whorf-Sapir Hypothesis, 134
Two-Switch Model, 85 word order, 97-8, 117, 133
types
- of bilinguals, 16-18, 46-8, ZISA Project, 313
118-19
Places, Languages and Peoples
(including linguistic minorities and minority
areas)
Africa(n), 157, 159, 160, 186, 188, Australia(n), 53-4, 160, 161, 202,
193, 195, 196, 237 242
Afrilcaan(s), 124-5, 208 Austria(n), 158, 159, 176, 183,
Albania(n), 220,222 184-5, 186, 190-1, 199, 201,
Allemanic, 171, 223, 250 208, 219, 222, 229-30, 266
Alsace (Alsace-Lorraine), Alsatian,
9, 48,51, 110, 171, 190, 200, Bahasa-Indonesia, 162, 218
202, 215, 220, 222-3, 225-6, Bahasa-Malaysia, 162, 218
230, 234, 240, 248-59, 267, 274 Balearic Islands, 261, 264, 266
Altbelgien (Old Belgium), 222, 227, Baltic republics, 54, 291
230 Barcelona, 261, 262, 264, 268, 271,
America(n), 7, 159, 161, 187, 193, 274, 279, 281, 283, 285-6, 288
195, 199, 202, 209, 235, 242, 266 The Basque Country (= Euskadi),
American Indian (Amerindian) Basque (= Euskera), 3, 42, 131,
languages, 28, 168, 187,209 162, 171, 190,203,209,211-12,
Amish, 160 216, 220, 222, 225, 229-30, 232,
Anatolia, 301 234, 237, 239, 260, 262, 263,
Andalusia!n), 271, 297, 301 265, 269, 287
Andorra, 219,227,261,264 Bas-Rhin, 250, 255
Aosta, 222 Belgium, Belgians, 2, 3, 10, 13, 14,
Arabic, 25, 106, 158, 167, 168, 208, 17, 34, 40, 158, 159, 160-1, 165,
218, 310 168, 173, 174, 176, 185, 187,
Aragon, 266, 271 188, 201, 208, 216, 219-23,227,
Armenia(n), 54, 208, 236 230, 231-2, 241-3, 246
Asia(n), 1, 157, 159, 160, 187, 188, Belorussia, 54
193, 196, 207, 209, 217, 237, 293 Bengali, 16
PLACES, LANGUAGES AND PEOPLES
Breton - see Brittany Czechoslovakia, Czech, 201, 229—30

Britain, UK, British, Britons, 2, 4,
14-15,34,40,44,116,121, ' Denmark, Danish, 2, 4, 17, 43,
142-3, 147, 153, 157-9, 161, 90-1, 158, 162, 172-3,200,
181, 198, 199, 201, 202, 209, 219- 26, 243, 289-90, 292
212-14, 219, 223, 225, 231, 233, Dutch, 3,52,71,77-8, 144, 158,
236, 240, 242-3, 307 159, 174, 226, 242
Brittany, Breton, 1, 44, 162, 187,
190, 199, 220, 222, 225, 232, England, English, 2, 3, 9, 16, 22, 25,
237, 239, 241, 267, 287 28, 39,40,44, 47, 48, 50,51,52,
Brussels, 16, 150, 153, 195, 200, 54, 55, 56, 57, 59, 60, 62, 66-7,
238, 267 68,71,73, 77-8,81-2, 82-3,
Bulgarian, 50 84-5,89, 90,91,93,96, 97,98,
99-100, 101, 102-3, 104, 105,
Cameroon, 188, 207, 209 106, 107, 111, 113, 114-15, 116,
Canada, Canadians, 4, 13, 16, 40, 121, 122, 124-5, 138, 141-6,
42, 47, 118, 123, 125-6,131, 151, 153, 158-9, 160-2, 171, 172,
140, 143-4, 145, 160, 165, 177, 181, 186, 187, 188, 189, 196, 201,
185, 201, 202, 210-11, 212, 217, 212-13, 218, 235, 242, 267
239, 242, 284, 287, 288 Estonia(n), 51, 52, 54, 57, 68, 107,
Cantonese, 4 110, 116, 195, 201, 208, 229,
Castile, Castilian Spanish, 154, 158, 234, 291
162, 171, 182, 189-90, 211-12, Ethiopia, 208
232, 246, 265-9, 272-3, 274-83, Europe(an), 1, 2-4, 9, 14, 21, 25,
287, 288 34, 40, 41,43^, 54, 101, 136,
Catalonia, Catalan, 3, 9, 17, 39^0, 139-41, 153, 158-74, 189,
42,44, 52,64, 110, 111-13, 118, 190-1, 193, 198, 199, 203, 215,
131, 154, 162, 164, 165, 171, 218, 219-24, 242
n182, 187, 188, 189, 190, 198, European Community, 54, 163, 215,
199, 202, 203, 204, 207, 208, 231, 233, 245, 307-8, 314-15
209, 211-12, 220, 222, 225-7, Euskadi, Euskera - see Basque
229-30, 232, 234, 240, 246,
260-88 Faroe Islands, Earoese, 190-1,
Celtic, 187, 190-1 220- 1, 223
Chicano, 50, 104, 115 Finland, Finnish, Finns, 2, 13, 14,
China, Chinese, 2, 50, 161, 196, 206 28, 51, 60, 127, 165, 188, 190,
Cornwall, Cornish, 186,187,220, 201, 216, 217, 219-21, 225, 229
224 Flanders, Flemish, Flemings, 3, 17,
Corsica(n), 162, 220, 226, 234, 237, 158, 160, 165, 174, 201, 208,
287 220, 223, 226-7, 232, 240, 248
Creole, 39, 116, 209, 240 France, French, 2, 3, 17, 34, 40, 42,
Croatian, Croat, 200, 220, 222, 292, 44, 48, 50, 51, 52, 54, 57, 60, 62,
300 69, 80, 98, 101-2, 105, 108, 111,
Cyprus, Cypriot, 231, 236 141, 157, 159-63, 164, 165, 167,
PLACES, LANGUAGES AND PEOPLES 347
168, 169, 170-1, 174, 188, 190, Haiti, Haitian Creole, 39, 167
199, 200, 202, 203, 204, 208, Haut-Rhin, 250, 255
213, 214,218,219-28, 230, Hawaii, 215
233^, 240-4, 248-59, 261-2, Hebrew, 52, 125, 151, 162,218
267 Hindi, 2, 161
French Canadian, Canadian French, Hiri Motu, 208
4, 16, 47, 123, 125-6, 131, Hispanos (Hispanic Americans), 34,
143^, 145, 165, 185, 201,202, 104, 109-10, 217
210-11, 217, 240 Hong Kong, 4, 104, 165, 206
Frisian, Frisian Islands, 2, 42, 165, Hungary, Hungarian, 51, 70, 176,
183, 187, 190-1,217, 220, 222, 184-5, 186, 190-1, 200, 201,
225-6, 234, 289-90, 292 204, 222, 229
Friulian, 220,222,224
Iceland(ic), 1, 197, 219
Gaelic, Gaels, 44, 121, 162, 176, India, Indian languages, 2, 13, 161,
187, 189, 190-1, 195, 196, 209, 165, 169, 199, 213, 216
213, 220, 223, 225, 227 Indonesia, 218
Gaeltacht, 250 Inuit, 4, 145
Galicia(n), 3, 44, 162, 171, 203, Inuktitut, 145
209, 211-12, 220, 222, 232, 234, Ireland, Irish Republic (Eire), Irish
260, 263, 265, 271, 297, 301 Gaelic, 121, 125, 187, 191, 195,
Garo, 50, 57, 66-7, 89 197, 209, 213-14, 219, 223-4,
Georgia(n), 51, 208 227, 234, 239, 250
Germany (also Federal Republic of Israel, 125, 162, 218
Germany), German, 2, 3, 4, 9, Italy, Italian, 2, 16, 28-9, 44, 51, 52,
10, 17, 34, 43,44, 48, 50, 51, 52, 57, 63, 71, 97, 101, 142-3, 153,
54, 55, 57, 59, 62, 63, 66, 67, 69, 159, 165, 170, 173-4, 181-2,
70-1, 80, 82-3, 90, 93, 97, 98, 186, 189, 198, 208,209,214,
99-100, 101, 103, 105, 106, 108, 218, 219, 223-6, 229, 232, 234,
111, 115-16, 118-19, 131, 138, 236, 240, 243, 290, 292-3, 295,
144, 147, 158, 159-60, 164, 165, 298, 300-2, 305, 310, 313
167, 168, 170-1, 172, 178, 184,
189, 190-1, 198, 199-200, 202, Jamaican Creole, 116
203, 208, 214, 217, 218, 219-30, Japanese, 17,51
234, 240 Jewish people, 176, 236, 242
Ghana, 207 Jurassim, Jurassian, 220, 223, 237
Girona, 261,264, 274, 285-6
Graubiinden, Grisson, 170,189,241 Kenya, 161,216
Greece, Greek, 97, 158, 167, 209, Kurdish, Kurds, 234, 292, 300
215, 220, 222, 236, 243, 290,
293, 295, 297, 298, 300-1, 310 Ladin, 220, 222, 224
Greenland, 2, 162,219 Lapland, Lapps (see also Same, 190,
Guarani, 168, 179, 180 220
PLACES, LANGUAGES AND PEOPLES
348
Latin, 16, 101, 158,200,203,208, Portugal, Portuguese, 1, 16, 158,

209,228,251,265,266 159, 197, 219, 243, 265, 290,
Latvia(n), 51, 58, 201, 204, 229, 293, 302, 313
234, 291 Provencal, 222
Lithuania(n), 201, 229, 234, 291 Prussia, 3, 199-200, 242
Lleida, 261, 264, 274, 285—6 Puerto Rico, Puerto Ricans, 39, 113,
Luxemburg, Luxembourgish 116,176
(Letzebuergesch, Letzeburgish),
13, 171,202,219-23, 230, 251, Quebec, 4, 201, 210, 239
253 Quechua, 209, 216
Macedodia(n), 215, 292, 301 Rhaeto-Romansch (= Romansch),

Magyars, 200, 220, 222 160, 165, 170, 186, 187, 189,
Malaysia, 207 206, 220, 234, 240-1, 250, 222-4
Manx. Isle of Man, 187, 220, 223, Romaguol, 222^
224 Romance languages, 66, 265
Mexico, Mexican, 90, 104, 202 Romania(n), 200, 290-2
Moldavian, 204 Russia(n), 2, 19-20, 42, 50, 51, 164,
Montreal, 123, 125-6, 177 169, 188, 218, 228, 229, 236, 292
Morocco, Morrocan, 144, 199, 243^ Ruthenian, 200
Neo Melanesian, 208 Samao, 207

Netherlands, 2, 34, 40, 47, 131, 144, Same (= Sami), 190,217,221,225
158, 161, 165, 173, 183, 217, Sard(inian), 220, 222, 234
219, 222, 225, 242-3, 289 Scandinavia(n), 4, 44, 118, 127, 161,
New Guinea, 176 172-3, 190, 217, 225
New Mexico, 206,215 Schleswig-Holstein, 190, 200, 222,
Nigeria, 160, 207 224-6, 289-90
N(ây, Norwegian, 7, 97, 190, Scotland, Scots, 44, 121, 176, 183,
199, 217,219 187, 189, 190-1, 199, 209, 213,
220, 223, 234, 267
Occitan, 220, 222, 225 Serbian, Serbs, 50, 292
Serbo-Croat(ian), 51,70
Pakistan(i), 160, 199 Sicily, Sicilian, 300, 301
Papua New Guinea, 208, 216 Singapore, 165, 216
Paraguay(ans), 13, 168, 176, 179, Slovak, 193,200,204
180 Slovene, 176, 183, 190-1, 200, 220,
Pennsylvania Dutch (Amish), 160 222
Persia, 218 Somali(a), 208
Peru, 209, 216 Serbian Sorbs, 313
Philippines, 207, 215, 216 South Africa, 124—5, 165, 208, 216
Piedmont(ese), 220, 222, 224 Soviet Union - see USSR
Poland, Polish, Poles, 4, 199-200, Spain, Spanish, 2, 3, 4, 17, 39^10,
201, 228, 229-30, 242, 290-2 42, 44, 51, 52, 55, 56, 59, 61, 64,
PLACES, LANGUAGES AND PEOPLES 349
67,81-3, 84, 89,90,91-2, 96, 295-8, 300-1, 306, 310,313

98, 99, 100, 101-2, 103, 104, South Tyrol (Tyrolean), 42, 159,
106, 110,111-13, 114-15, 147, 182, 220, 222, 225-6, 232, 237
157-9, 161, 164, 168, 171-2,
179, 180, 187, 198-9, 202, 203, Uganda, 208
204, 209, 210, 211-12, 214, 216, Ukraine, Ukrainian, 4, 54, 125, 204,
217, 218, 219-22, 225-7, 229, 208, 240, 242
232, 234, 240, 243^, 260-88, Urdu, 4, 240
290, 293, 297-8, 300, 302, 305, USA (United States of America), 4,
310, 313 16, 34, 39, 40, 48, 104, 109-10,
Strasbourg, 190 114-15, 116, 118, 120, 121, 125,
Sutherland, 176, 189 141-3, 145, 160, 161, 176, 189,
Swahili, 42, 161, 209 209, 210, 215, 217, 236, 237,
Sweden, Swedish, 28, 39, 44, 51, 242-3
54, 57, 58, 83, 85, 110, 116, 119, USSR (Soviet Union), 2, 42, 43, 54,
127, 188, 190, 217, 219-21, 225, 161, 188, 199, 206, 208, 234,
234, 243 241, 290-92
Swiss German (Schwyzerdiitsch,
Schwyzertuiitsch), 167, 168, 170, Valdotains, 222
241, 248, 250, 253 Valencia, Pais Valencia, 226, 260-1,
Switzerland, Swiss, 2, 10, 13, 14, 264
160, 165, 170-1, 173, 186, 189,
206, 208, 216, 219-21, 223, 232, Wales, Welsh, 3-4, 14, 16, 42, 43,
234, 236, 241, 243^, 248, 44, 91, 118, 121-2, 131, 153-^,
250-1, 253 162, 165, 179, 186, 187, 188,
189, 196, 198, 199, 201, 202,
Tagalog, 216 207, 208, 209,212,213,214,
Taiwan, 206 220, 223, 225-6, 234, 237, 239,
Tanzania, 42, 161, 169, 209 240-1, 267, 284, 288
Tarragona, 261, 264, 274, 285-6 Wallonia, Walloons, 165,182,201,
Ticino, Ticinese, 170, 173^, 220, 220, 222, 227, 232,/234
223 West African Pidgin English, 209
Tok Pisin, 208 West Germany - see Germany
Tunisia, 208
Turkey, Turkish, 4, 17, 24, 34, 54, Yiddish, 234, 236
131, 144, 173, 201, 206, 209, 218, Yugoslavia(n), 2, 197, 243, 290,
231, 234, 236, 243-4, 290, 292-3, 292-3, 295, 298, 301, 310, 313
Author Index
Abeya, 280 Bourhis, 181

Adler, 18, 19, 33, 35 Brown, 72-3, 106, 117
Aellen, 144-5 Bruner, 91
Aguirre, 110, 114 Burling, 50, 57, 60, 66-7, 89, 150
Albert, 20 Byram, 30, 43, 141
Allardt, 247, 263
Appel, 9, 104, no, 115, 144 Calsamiglia, 110, 111-12, 116
Amau, 282, 283, 284 Carr, 190
Amberg, 40, 51, 76, 83, 84-5, 108 Carrow, 70
Arsenian, 122 Celce-Murcia, 51
Auer, 116,313 Cellard, 254
Auemheimer, 301, 312 Champagne, 65-6
Azevedo, 273, 277, 288 Child, 142-4, 146
Chomsky, 24, 36
Baetens Beardsmore, 18,20,21,23, Christopherson, 21
24-5,31,32,43, 146, 174, Churchill, 141,242,247
176-7, 267, 273 Clahsen, 302,304,313
Bain, 126 Clyne, 80, 304, 313
Baker, 120-1, 122-4, 126, 132-3, Cobarrubias, 207,215-16
138, 141, 154 Cole, 257-8
Bausch, 313 Cornejo, 113
Ben-Zeev, 125, 126 Crystal, 2
Berber, 298 Cummins, 118, 119, 125, 126,
Bergman, 76 127-31, 133, 141
Berlin, 37 Cunze, 52
Berman, 151
Blan, 9, 10 Darcy, 122
Blom, 184 De Avila, 120
Bloomfield, 15,21,22,28 De Cicco, 269
Boada, 282, 283, 284 De Houwer, 52, 68, 69, 71-2, 76,
Boos-Niinning, 141 77-8
Borel-Maisonny, 141 De Jong, 40
Borris, 297 Desgranges, 49, 54, 312, 313
AUTHOR INDEX 351
De Villiers, 61 Giles, 181

Diebold, 21 Giner, 262, 268
Diller, 20 Goldberg, 142
Di Luzio, 116,313 Greene, 190-1
Dittmar, 304-6, 313 Greenfield, 176, 178
Dodson, 92 Grosjean, 39, 59-60, 62, 74—5, 76,
Dorian, 176, 183, 189, 190-1 84, 94, 96, 102, 110-11, 114,
Doyle, 65-6 133, 140, 151,153
D’Oyley, 141 Gumperz, 88-9, 104, 110, 176, 184,
Durkheim, 146 190-1
Edwards, 187, 191, 193-6, 197,213, Halliday, 21, 91

235, 236, 245, 247 Hamers, 9, 10
Elias-Olivares,104 HansegSrd, 28
Elwert, 48 Harding, 40
Emrich, 50 Harrison, 91
Ervin, 20 Harshman, 37
Ervin-Tripp, 88-9 Hatch, 51
Haug, 248
Fabre, 279 Haugen, 7, 18, 22
Fantini, 49, 52, 56, 80, 84, 89, 90, Havelka, 22
102, 106, 124 Heath, 213
Fasold, 5, 159, 168, 169, 177, 178, Heraud, 225
183 Herder, 119,203
Ferguson, 166-7, 169, 180, 205-6, Hem^dez-Chavez, 104, 110, 114
254 Hernandez Pina, 61
Ferrer i Girones, 226-7 Hewitt, 116
Fichte, 119,203 Hinderling, 247
Fishman, 89, 167-9, 176, 177-8, Hoffmann, 52, 55, 58-9, 60, 68,
179, 196-7, 202, 254, 266 82-3, 84, 106,
Foster, 247 Hooper, 270,271
Francis, 194 Houston, 18
Fritsche, 298, 313 Hoyer, 50
Humboldt, 136, 153, 203-4
Gal, 176, 183, 184, 190-1 Husen, 141
Gardner, 22, 144—5, 245
Gardner-Chloros, 110,250,254, lanco-Worrall, 124—5,126
256, 257 Imedadze, 51, 70, 76, 84, 124
Garvin, 202 Isajiw, 194
Gauthier, 256-7 Itoh, 51
Geiselberger, 293
Genesee, 76, 78, 105, 107, 140 Jacobovits, 20
Gerth, 200 Jahn, 136, 138
Gibbons, 104 Jakobson, 56
352 AUTHOR INDEX
Jemudd, 207 Mathiot, 202

Jonekeit, 26, 40 McClure, 113, 116
Jones, 122-3 McIntosh, 21
J0rgensen, 172-3 McLaughlin, 18, 37, 48, 58, 59, 67,
Just, 146 76,104, no
Meisel, 49, 51, 54, 69, 76-7, 304,
Kedourie, 197 306, 313
Keilhofer, 26, 40 Metraux, 60, 84
Keim, 305-6, 313 Mikes, 51, 70, 71
Kelly, 23 Miller, 147
Kessler, 70, 81-2, 91-2, 149-50 Mills, 200
Klein, 304-6,313 Mkilifi, 42, 110
Kloss, 205, 207-9 Murrell, 51, 55, 60, 67, 150
Krashen, 37 Muysken, 9, 104, 110, 115, 144
Kutsch, 49, 54, 312, 313
Kwan, 196 Nelde, 7, 185, 313
Nelson, 62
Labov, 258 Neustupny, 207
Ladin, 255-6, 258, 259 Neville, 256-7
Lado, 7, 97 Nikolinakos, 302-3
Lambert, 20, 21, 22, 118, 119,
123-4, 143-5, 146 Obler, 20
Lapkin, 140 ^ Oestreicher, 21
Laurie, 136, 138 Oksaar, 23, 29, 49, 51, 54, 55, 57,
Lebmn, 141-2 74, 76, no, 116
Leman, 141 Oktem, 53
Lenneberg, 36 Oomen-Welke, 311
Leopold, 48, 50, 56-7, 62, 66, 70, Opper, 141
124 Orlovic-Schwarzwald, 305,313
Lepschy, 234 Osgood, 20
Lewis, 42, 43
Lieberson, 177 Padilla, 76, 113
Liebman, 76, 113 Paradis, 20, 141-2
Lindholm, 76,113 Parasher, 178, 180
Park, 49, 51, 54, 76, 105, 106
Macaulay, 213 Paulston, 28
Mackey, 13, 15-16, 23, 24, 95, 96, Pavlovitch, 50
102, 209-10 Peal, 123-4
MacNab, 123-4^ Penfield, 36, 85
Macnamara, 20, 22, 85 Peters, 61-2
Mandabach, 213 PfaflF, no, 118, 313
Maring, 269 Philipps, 248
Marshall, 217 Pichon, 136, 141
Martin-Jones, 133 Pienemann, 313
AUTHOR INDEX 353
Piette, 91 S0nderg4rd, 150-1

Pohl, 165, 166 Spolski, 140
Poplack, 104, 110, 113, 114-15, 116 Stavens, 52
Porsche, 153 Stephens, 200, 221, 229, 230, 247,
Portz, 313 248, 251, 269
Power, 246 Stolting, 133, 313
Stmbell i Trueta, 269, 271, 286
Raffler-Engel, 51,57,70 Subirats, 275
Rath, 313 Swain, 51, 76, 84, 140
Redlinger, 49, 51, 54, 76, 105, 106
Rehbein, 53,313 Tabouret-Keller, 48, 51, 256, 257
Reyes, 104, 115 Taeschner, 49, 51, 52, 62, 63^,
Riley, 40 70-1, 72, 76, 79, 106
Ris, 170 Taylor, 181
Roberts, 36, 77, 85 Timm, 104, 114, 190
Romaine, 9, 74, 98, 133 Tosi, 142-3, 144, 146
Ronjat, 50, 57, 80 Toukomaa, 28, 127, 130
Rubin, 176, 179, 180, 182-3, 207 Travis, 141
RuKe-Draviqa, 51, 58, 60 Tmdgill, 258
Tucker, 118, 144-5
Sabater, 275 Tuson, 110, 111-12, 116
Saer, 122
Sander, 136 Valdes Fallis, 104, 110, 111
Sankoff, G, 176, 183 Vallverdu, 270
Saunders, 23, 40, 49, 51, 54, 62, 76, Verdoodt, 168
80, 97-8, 105 Vihman, 52, 59, 68, 106-7, 108
Schlapfer, 206 Vila, 52, 64, 71
Schlyter, 105, 106, 108, 117 Volterra, 51, 62, 63, 75-6, 79
Schmidt-Rohr, 136, 138, 142, 178 Vygotsky, 124
Schumann, 303, 313
Scott, 125-6 Walker, 190-1,289-90
Segalowitz, 65-6, 146 Wandruszka, 10
Seliger, 37 Wardhaugh, 2, 5, 217
Shapston, 141 Weber, 194, 196, 200
Shibutani, 196 Weinreich, 7, 14, 15, 19-20, 23, 77,
Siguan, 283 95, 138, 142
Silva-Cprvalan, 104, 110, 112, 115 Weisgerber, 136, 138, 142
Skinner, 36 Wells, 73
Skutnabb-Kangas, 8, 18, 19, 20, 23, Wesche, 84
25,26-27, 28,46-8, 118, 119, Williams, 237-8, 239, 245
127, 128, 130, 141, 173 Wolff, 298
Smith, A, 196,238 Wolk, 20-1
Smith, M E, 50 Woolard, 269
Soffieti, 29
LoNCiMAN* Linguistics* Library
General Editors
R. H. Robins, University of London
Martin Harris, University of Essex
During the last decade considerable advances have been made in

our understanding of bilingualism. This progress has made a fresh
introduction to the subject all the more timely. An Introduction to
Bilingualism provides a comprehensive review of the most important
aspects of individual and societal bilingualism and evaluates the
current state of relevant research.
This innovative and accessible survey examines both theoretical and

practical issues. At the level of the individual it addresses such
questions as: What is involved in the study of bilingual children?
What are the patterns of bilingual language acquisition? In which
ways do the language competence and the speech of bilinguals
differ from those of monolinguals? Topics that sometimes arouse
controversy are explored - such as the question of whether there is
a relationship between bilingualism and a child’s cognitive,
psychological and social development.
The book is also concerned with multilingualism, that is,

bilingualism as a societal phenomenon. It focuses on such issues as
language choice in bilingual and multilingual communities, national
identity and the education of bilinguals. The inclusion of several
case studies of European linguistic minorities serves to exemplify
the topics dealt with at the theoretical level and to illustrate the
linguistic complexities found in contemporary Europe.
Assuming no prior knowledge of the subject, this volume will prove

invaluable to students of linguistics, sociolinguistics and
psycholinguistics, as well as being welcomed by teachers, parents,
social workers, educational administrators and all those who come
into contact with bilingual children and adults.
Charlotte Hoffmann is Lecturer in German and Linguistics at the

University of Salford.
Front cover illustration by Chris Corr.
ISBN D-saa-a^mB-?
PI PI PI
Longman 9 78Q 582 291430

An Introduction To Bilingualism

Uploaded by

Copyright:

Available Formats

An Introduction To Bilingualism

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

An Introduction To Bilingualism

Uploaded by

Copyright:

Available Formats

ftnHnrr tVift Innnrii^^fTn P'

i uui iiwHim" > i. C l

•m vi.bru^r vor s cgnc

"Als sicli^ui hierfioh iiaba- rrdch

ijiic uno pcx

General editors Pidgin and Creole Languages

Structural Aspects of Language A History of English Phonology

Published in the United States of America

© Longman Group UK Limited 1991

All rights reserved; no part of this publication may be

First published 1991

British Library Cataioguing in Pubiication Data

Library of Congress Cataioging in Pubiication Data

Set in Linotron 200 10/12pt Times

Produced by Longman Singapore Publishers (Pte) Ltd

PART I PSYCHOLINGUISTIC ASPECTS OF BILINGUALISM

2 The study of bilingual children 33

3 Patterns of bilingual language acquisition 55

4 Aspects of bilingual competence 74

5 Features of bilingual speech 94

5.5 Code-switching 109

6 Cognitive and educational aspects of bilingualism 118

6.1.2 Overview of research on bilingualism and cognitive

7 Sociocultural aspects of bilingualism 136

PART II SOCIOLINGUISTIC ASPECTS OF BILINGUALISM

8 Societal multilingualism 157

8.2.1 Historical factors 158

9 Language choice, language maintenance and

10 Language and national identity 193

11 Linguistic minorities 219

11.2 Two main types of indigenous linguistic minorities 224

12 Case Study I: The Alsatians 248

13 Case Study II: The Catalans 260

14 Case Study III: Migrant workers in the Federal Republic of

14.2.3 Reasons for migration and problems

Appendix: Council of the European Communities

Subject Index 339

Index of Places, Languages and Peoples 345

Author Index 350

11.1 Indigenous linguistic minorities in western Europe 220

During the last decade or so researchers into bilingualism have been

courses on Bilingualism taught to final-year Modem Language stu¬

undertake research in the context of linguistic minority settings. Many

The Publishers are grateful to the following for permission to repro¬

Sociology and Social Psychology, University of Tampere for a figure

NOTE: All references to the Federal Republic of Germany in this

This book is for

Paco, mi bilingiie favorito;

fiir meine Kinder Cristina und Pascual;

og til min mor, som startede det hele.

Bilingualism - a natural phenomenon

The role of attitudes

Bilingualism as a problem and as a resource

example, Fasold (1984: 4) suggests that ‘multilingualism works against

Any such process is likely to lead to tension and resentment on the

negative attitudes towards bilingualism). A second possible advantage

The remit of the study of bilingualism

(1) the individual;

A wide range of social sciences contribute towards our under¬

‘ , 4 ')i_ «1^ -O£r';r'0-Jp^ ’itm