Vous êtes sur la page 1sur 240

cover next page >

title: Papers From the Fifth Nordic Conference On Bilingualism


Multilingual Matters (Series) ; 41
author: Gimbel, Jørgen.
publisher: Multilingual Matters
isbn10 | asin: 1853590142
print isbn13: 9781853590146
ebook isbn13: 9780585224800
language: English
subject Bilingualism--Congresses, Education, Bilingual--Congresses,
Second language acquisition--Congresses, Language and
education--Congresses.
publication date: 1988
lcc: P115.N67 1987eb
ddc: 404/.2
subject: Bilingualism--Congresses, Education, Bilingual--Congresses,
Second language acquisition--Congresses, Language and
education--Congresses.

cover next page >


< previous page page_i next page >
Page i

Papers from the Fifth Nordic Conference on Bilingualism

Edited by
J. Gimbel, E. Hansen,
A. Holmen and J. N. Jørgensen
MULTILINGUAL MATTERS 41
Series Editor: Derrick Sharp

MULTILINGUAL MATTERS LTD


ClevedonPhiladelphia

< previous page page_i next page >


< previous page page_ii next page >
Page ii
Library of Congress Cataloging in Publication Data
Nordiska tvåspråkighetssymposiet (5th: 1987:
Royal Danish School of Educational Studies)
Papers from the Fifth Nordic Conference on
Bilingualism.
(Multilingual Matters: 41)
Held at the Royal Danish School of Educational
Studies at Copenhagen, June 22-25, 1987.
Bibliography: p.
Includes index.
I. BilingualismCongresses. 2. Education,
BilingualCongresses. 3. Second language acquisition
Congresses. 4. Language and educationCongresses.
I. Gimbel, Jørgen. II. Title. III. Series.
P115.N67 1987 404'.2 88-10031
British Library Cataloguing in Publication Data
Nordic Conference on Bilingualism (5th: 1987:
Copenhagen, Denmark).
Papers from the Fifth Nordic Conference
on Bilingualism. (Multilingual Matters : 41).
1. Bilingualism
I. Title II. Gimbel, J. (Jurgen)
404'.2
ISBN 1-85359-014-2
Multilingual Matters Ltd
Bank House, 8a Hill 242 Cherry Street
Road & Philadelphia, PA 19106-
Clevedon, Avon BS21 7HH 1906
England USA

Copyright © 1988 J. Gimbel, E. Hansen, A. Holmen, J. H. Jørgensen and the


authors of individual chapters.
All rights reserved. No part of this work
may be reproduced in any form or by any means
without permission in writing from the publisher.
Typeset by Photo-graphics, Honiton, Devon
Printed and bound in Great Britain by Short Run Press.

< previous page page_ii next page >


< previous page page_iii next page >
Page iii

Contents
Preface
Jørgen Gimbel, Elisabeth Hansen, Anne Holmen And Norman
Jørgensen 1
Introductory Note
Jørgen Gimbel, Elisabeth Hansen, Anne Holmen And Norman
Jørgensen 3
Opening Address
Bertel Haarder 7
Bilingualism in Education and Bilingual Education:
The State of the Art in the United States
Anna Uhl Chamot 11
Bilingual Education and Bilingualism in Education:
A Comment
Gabriele Kasper 37
The Cognitive Academic Language Learning Approach (CALLA)
J. Michael O'malley 43
Social and Intellectual Functions of Language:
A Fruitful Distinction?
Hans Vejleskov 61
Lexical Characteristics of Near-Native Second-Language Learners of
Swedish
Kenneth Hyltenstam 67
Syntax and Information Structure in Learner Language
Anne Holmen 85
Is There Any Order? On Word Order in Swedish Learner Language
Language Acquisition
Maria Bolander 97
Some Psychological Aspects Of Early Second Language Acquisition
Mirjana Vilke 115
Immigrant Children's SwedishA New Variety?
Ulla-Britt Kotsinas 129
Language Varieties and Intercultural Education
Sigrid Luchtenberg 141
Supplementary Mother-Tongue Education and the Linguistic
Development of Yugoslav Children in Denmark
Andrina Pavlinc-Wolf *, Karmen Brcic* And Nadezda Jeftic* 151
Bilingual Education in Yugoslavia: Some Experiences in the field of
Education for National Minorities/nationalities in Yugoslavia
Sonja Novak-Lukanovic* 169

< previous page page_iii next page >


< previous page page_iv next page >
Page iv
Language Policy and Language Planning after the Establishment of the
Home Rule in Greenland
Aquigssiaq Møller 177
Linguistic Minorities and Language EducationThe English Experience
Euen Reid 181
Latest Developments in Early Bilingual Education in France and
Southern Europe
Klaus-Erich Gerth 193
Bilingualism, Education and Identity
John Edwards 203
Where Do We Go From Here?
Concluding Panel Discussion 211
Appendix. Contents of Companion Volumes 229
Index 233

< previous page page_iv next page >


< previous page page_1 next page >
Page 1

Preface
This volume is one of three publications of proceedings from the Fifth Nordic Conference on Bilingualism. The remainder of the
papers are published by Multilingual Matters Ltd. in the Copenhagen Studies in Bilingualism, volumes Four and Five. In
Volume Four the papers on Bilingualism and the Individual are grouped together. They include the linguistic papers, dealing
with either interactional analysis, linguistic structure, or variation. Here are also the papers on different backgrounds and their
effects on second language learning, and finally the papers, small in number, but substantial in content, about literature and
bilingualism.
Volume Five of the Copenhagen Studies comprises the papers on Bilingualism in Society and School. There are papers on
language planning, on status differences between languages, and on language ecology. This volume also includes papers on
group bilingualism, papers on pre-school children, and finally papers on classroom practice. The contents lists of the two
companion volumes appears in the Appendix.
Some of the exhibits from the shows and exhibitions at the conference are used as illustrations in these volumes.
Here, we as conference organisers, wish to express our gratitude to:
our co-organisers, especially Marie Hald and Inger Nørgaard, our colleague Bent Søndergaard of the Pedagogische Hochschule
Flensburg, and, from our own institute, Gerd Gabrielsen, Erik Larsen, Jørn Lund, and Marie-Alice Séférian;
the participants in the conference;
and our sponsors:
King Frederik and Queen Ingrid's Foundation
Queen Margrethe and Prince Henrik's Foundation
The TUBORG Foundation
Knud Højgaard's Foundation
Rosalie Petersen's Foundation
Julius Skrike's Foundation
SAS Travel Grants
The Rockwool Foundation
The Foundation for Danish-Swedish Cooperation
The Cultural Foundation for Denmark and Finland
Denmark's Teachers' Union
The Danish Masons
The Letterstedt Society
The Ministry of Education
The City of Copenhagen

< previous page page_1 next page >


< previous page page_2 next page >
Page 2
The County of Copenhagen
The Research Council for the Humanities
The Greenland Foundation of the Danish Parliament
The Royal Danish School of Educational Studies
JØRGEN GIMBEL
ELISABETH HANSEN
ANNE HOLMEN
NORMANN JØRGENSEN

< previous page page_2 next page >


< previous page page_3 next page >
Page 3

Introductory Note
Jørgen Gimbel, Elisabeth Hansen, Anne Holmen and J. Normann Jørgensen
Department of Danish Language and Literature, The Royal Danish
School of Educational Studies, Emdrupborg,
DK-2400 Copenhagen NV, Denmark
The Fifth Nordic Conference on Bilingualism was held at the Royal Danish School of Educational Studies at Copenhagen, 22-25
June 1987. The main theme of the conference was Two Languages and Two Cultures in Education. This theme was dealt with in
many ways by the contributions offered during the conference, and no less in the more than 1,300 pages of manuscripts turned
in for publication in the proceedings. A large number of the papers, however, deal with certain aspects of bilingualism, currently
drawing interest from many students of the discipline. These aspects include the macro-level planning of language education in
multilingual societies, the micro-level second language learning strategies, the linguistic level analysis of interactional data
involving bilinguals, and the cultural level of bilingual writers.
As our editorial principle we have chosen to group the papers according to their relation to the main theme of the conference.
Two of the sub-themes are also presented in the present volume. We have, on the one hand, the papers dealing with the schools'
attitudes to bilingualism, discussed at the national level and, on the other hand, papers about the learning and development
processes which take place in the bilingual individual.
One reason for grouping these two aspects together for this purpose is the fact that they have been particularly controversial in
Scandinavia, for example, Paulston (1983) in the case of Sweden, and Jørgensen (1985) in the case of Denmark, both testify to
the fact that the question of the organisation of classes for linguistic minority children in schools has been hotly debated in these
countries. This was also clearly reflected in several papers at the third and fourth conferences in this series (see, e.g. von Essen,
Löfgren and Ouvinen-Birgerstam and Søndergaard (In Ejerhed and Henrysson, 1981). See also Wande et al., 1987, e.g. Wande
and Martin-Jones and Romaine).

< previous page page_3 next page >


< previous page page_4 next page >
Page 4
The discussion about organisational principles is briefly touched upon by the Danish Minister of Education, Bertel Haarder, in
his opening address in which he states that the organisation of education for the linguistic minority children in Denmark has not
yet 'reached any final or near-final stage'.
The question of cognition and communication as two separate linguistic functions is also well known in the tradition of these
conferences. In 1980 Robbins Burling challenged the notion of semi-lingualism, as it was used in Swedish educational circles,
with reference to the ups and downs of Black English in U.S. educational history (Burling in Ejerhed and Henrysson, 1981).
Four years later, Martin-Jones and Romaine severely criticised the distinction between basic interpersonal communication skills
(BICS) and cognitive/academic language proficiency (CALP) as well as the concept of semi-lingualism (Martin-Jones &
Romaine in Wande et al., 1987). They argue that language cannot be compartmentalised as simply as the hypothesis suggests,
nor can the so-called thresholds be described in terms of observable linguistic or educational data or events.
These two themes are both represented in the main plenary lecture given by Anna Uhl Chamot. Firstly, her paper describes the
multitude of bilingual education programmes in the United States of America, and secondly it goes on to review some examples
of the research currently being undertaken with respect to bilingualism and teaching in that nation, especially studies of learners'
strategies and the development of a language learning programme built on the notion of CALP.
Gabriele Kasper, in her comments on Chamot's lecture, focusses on the research into learning strategies reported by Chamot.
She challenges the assumption behind it, referring to, among others, Martin-Jones & Romaine (the 1986 version of their paper),
and calling for a comprehensive and explicit model of second language learning. This is once again taken up in the concluding
remarks by Kasper, in which she calls for extensive crossdisciplinary and multidisciplinary co-operation involving cognitive
psychology, anthropology, linguistics, etc., in second language acquisition research.
The paper by J. Michael O'Malley continues the discussion by presenting the CALLA (Cognitive Academic Language Learning
Approach) in some detail, thus attempting to make the BICS-CALP distinction operable, and the discussion also involves Hans
Vejleskov who refers to data from Norwegian pre-school children's role play to argue that the distinction between 'social' and
'intellectual' conversation cannot be maintained.
The following papers all present empirical linguistic data from second language learners to shed light on the processes involved
in language acquisition. Kenneth Hyltenstam's lexical data from near-native learners of Swedish are used to question the notion
of 'context-embedded' and 'cognitively undemanding' communication and to shed light on the concept of 'fossilisation'. Anne
Holmen and Maria Bolander both present data from analyses

< previous page page_4 next page >


< previous page page_5 next page >
Page 5
of the acquisition of syntax by immigrants. Holmen's data stem from an investigation of the acquisition of Danish by Vietnamese
and Albanian-speaking adolescents. Bolander's informants are Finnish, Polish and Spanish.
Mirjana Vilke's contribution is an overview of several projects in early foreign language learning, in this case English by
Serbocroatian speakers in Yugoslavia. The importance of motivation, and the sources of motivation, are emphasised, and it is
reported that young second language learners mainly seem to be at an advantage as far as pronunciation is concerned.
Ulla-Britt Kotsinas discusses the linguistic effects of the acquisition of Swedish by linguistic minority students in terms of
creolisation. Without coming to a firm conclusion she finds that the development of Swedish with the adolescents shows some
creolisation characteristics. Sigrid Luchtenberg also deals with the linguistic variation which she relates to cultural
developments, and consequently linguistic competence to sociocultural knowledge. This has obvious educational implications,
some of which are also discussed in her paper.
Pavlinic-Wolf *, Brcic* and Jeftic* present some of the data from the Danish part of a questionnaire project involving Yugoslav
children and adolescents in several Western European countries. The paper compares the students' assessments of their first and
second language skills. The possible determining factors in the development of the languages are discussed in view of the results
from the questionnaires. For example, it was found that while the students would often know the grammatical rules of their
mother tongue, they would fail to apply them in real-life communication, leading the authors to conclude that a 'passive'
knowledge of language cannot be readily brought to life.
The complexity of second language learning as it appears in these papers, and especially the multidimensional problems that
research has to grapple with, were again touched upon in the final remarks by Kasper, Andenæs, Hyltenstam, Reid and Chamot.
Their unanimous call for co-operation between scholars in different fields, between researchers and practitioners, between
teachers and educators in different fields of education and between nations, underlines the tendency of several of these papers.
The second sub-theme represented in this volume concerns the schools' attitudes to bilingualism. Novak-Lukanovic* outlines the
typology of bilingual education in Yugoslavia in which the teaching of minority languages is motivated by terms like 'the
language of national minorities becomes an appropriate instrument for communication in public and social life'. Aqigssiaq
Møller's lecture is here summarised by Gerd Gabrielsen. It presented the political and educational initiatives undertaken in
Greenland since the introduction of Home Rule in 1979. The schools are seen as an instrument to strengthen the Western
Greenlandic language, but the efforts are severely hampered by a serious shortage of teachers competent in that language (see
also Møller, 1988). Euan Reid describes the English scene, the composition

< previous page page_5 next page >


< previous page page_6 next page >
Page 6
of the minorities, and the different approaches to the task of educating their children. He favours the introduction of minority
languages into the school, and in teacher training, but predicts little success in this respect. The opposite is the case for Southern
Europe. Klaus-Erich Gerth presents an overview of current developments in Spain (Catalonia), Italy and France. In all three
countries, bilingual education is being promoted, and for younger and younger children.
These are four very different societies. We have Yugoslavia with its multitude of indigenous languages. We have Greenland
with an educational system in which the mother tongue of the majority is not the language of most teachers, a nation striving to
strengthen its national language. And we have Britain, especially England, in which the minority languages are underprivileged
in the schools, not the least because a good many people in minorities are immigrants. And furthermore, we have Italy and
Spain, countries which have experienced both internal migration and emigration, for example to England and they are countries
now, as it seems, actively promoting bilingualism. The overviews presented here all seem to accept the belief that the fate of
languages may be determined to at least a nonnegligible degree by educational policies. John Edwards challenges this belief in
his paper. He argues that schools can do little to support linguistic minorities with respect to their group identity, or the
maintenance of their languages. This is a challenge not only to the previous papers, but also to the belief underlying much of the
current Scandinavian discussion, as was also pointed out by Burling (in Ejerhed and Henrysson, 1981) who went so far as to
question whether the schools had the right to interfere at all. This is another question which will have to be answered elsewhere.

References
Ejerhed, E. and Henrysson, I. (eds (1981) Tvåspråkighet. Föredrag från tredje Nordiska Tvåspråkighetssymposiet 4-5 juni 1980,
Umeå universitet. Acta Universitatis Umensis. Umeå Studies in the Humanities 36, Umeå: Umeå universitet.
Jørgensen, J. Normann (1985) Skolefag, sprog eller slagmark. Artikler og foredrag om dansk som fremmedsprog i
undervisningen. Copenhagen Studies in Bilingualism, Vol. 1. (openhagen: The Royal Danish School of Educational Studies.
Martin-Jones, M. and Romaine, S. (1986) Semi-lingualisma Half-Baked Theory of Communicative Competence. Applied
Linguistics, Vol. 7, No. 1, 26-38.
Møller, A. (1988) Sprogpolitik og sprogplanlægning etter hjemmestyrets indførelse. In J. N. Jørgensen et al. (eds), Copenhagen
Studies in Bilingualism Vol. 5. Clevedon: Multilingual Matters Ltd.
Paulston, C. B. 1983) Swedish Research and Debate about Bilingualism. A critical review of the Swedish research and debate
about bilingualism and bilingual education in Sweden from an international perspective. Stockholm: National Swedish Board of
Education.
Wande, Erling, Anward, J., Nordberg, B. Steensland, L. and Thelander, M. (eds) (1987) Aspects of multilingualism. Proceedings
from the Fourth Nordic Symposium on Bilingualism 1984. Acta Universitatis Upsaliensis. Multiethnica Upsaliensia 2. Uppsala:
Uppsala Universitet.

< previous page page_6 next page >


< previous page page_7 next page >
Page 7

Opening Address
Bertel Haarder, MP
Minister of Education, Denmark
Dear Teachers, Administrators and Distinguished Scholars
Welcome to Denmark, a small nation, to some a remote corner of Europe, to us the centre of the world. It is particularly suitable
that an international conference on bilingualism in education takes place here now. For years Denmark's educational system has
been preoccupied with teaching foreign languages to its Danish-speaking students, and we have thought only a little about
bilingualism in education. We have (more or less) been aware of the fact that a comparatively small minority of students need to
be taught Danish as a second language (namely the Faroese, The Greenland and the Schleswig children), but this fact never had
any serious impact on the educational system as a whole. Some ten years ago, however, the influx of immigrant and refugee
families brought about a change in this. Our educational system began to move and did so in a typical Danish way. The schools
adjusted to the new situation by their own initiative. In several school districts, especially in metropolitan Copenhagen, teachers
worked hard, often with great imagination and dedication, to develop a pedagogy for the teaching of linguistic minority children.
This was in fact very much in line with Danish tradition. For many years the public schools of Denmark have developed from
the bottom up in a decentralised way. New methods were tested, new materials designed, new ways of organisation invented, all
of it happpening on the local level. The laws and regulations were written and issued after the new developments had become
reality in the schools. Most likely we have not yet reached any final or near-final stage in the adjustment of the Danish school to
the linguistic minority students. We see a wide range of new projects carried out by enthusiastic teachers and administrators,
which will more than likely lead to further changes in our national guidelines for the schools, issued by the Ministry.
In 1984 the Ministry of Education issued the first comprehensive regulatory document on the teaching of linguistic minority
students. In some ways it is already obsolete. Fortunately, I wish to add. I see it as a sign of good health that the schools are
steadily trying new ways, and I am happy to see that so many Danish educators have taken this opportunity to learn from their
colleagues abroad. An opportunity like this is also a chance to show that we, too, feel a responsibility towards international co-
operation. At the conference of European ministers of education in Dublin, 1983, a

< previous page page_7 next page >


< previous page page_8 next page >
Page 8
recommendation which especially calls for an intercultural dimension in education was adopted.
Since the linguistic minorities in Denmark constitute such a multitude of different languages and cultures, our resources are far
from sufficient to develop our own pedagogy for each group. We realise that we need to cooperate internationally, especially
with some of the donor countries. This gives us the pleasant burden of having to contribute our share to international research
and development. Gradually, therefore, the new facts of life in Denmark are being taken up and dealt with in research and
scholarship. We hope to be able to see this trend accelerate in the coming years.
Of course, bilingualism is not in itself something new in Denmark. Ludvig Holberg wrote in several languages, Hans Christian
Andersen knew a couple himself. And generations of grade school students have learnt one or more of the European languages,
English and German, to mention a couple of them. This fact should not be forgotten when we discuss the 'new' bilingual
students in our schools. The study of foreign language acquisition and foreign language learning is advancing in Denmark. This
field is continuously being developed. One example of this is the nationwide experiment with the introduction of French as the
second foreign language in the seventh grade in the public school. Other experiments deal with Spanish, Portugese and Japanese
in high school. There is good reason to believe that we have firm ground to build on there, with respect to the development of a
second language pedagogy for Danish. There is also little doubt that we can learn very much from the experience of The
Faroese Islands, Iceland, Schleswig and especially Greenland during the past eight years. In these places Danish has been taught
as a second language for years, in steadily more refined systems.
These three aspects of bilingualism in education are particularly important as seen from the point of view of the Danish Central
Administration. We are very concerned with the maintenance of successful bilingualism on all these three fronts. In a modern,
industrialised and, to a large extent, urban society, international co-operation and contact make individual bilingualism,
including a wide range of languages, necessary. It remains to be seen whether it becomes a reality, but we must acknowledge
that it is a possibility that multilingualism and multiculturalism will be characteristic of the Scandinavian countries, including
Denmark, in the future.
I should like to mention here that about 50 different languages are provided in the Danish folkeskole as mother tongue teaching
to immigrants and refugees. We hope that these children will contribute to the increased knowledge of their culture among
Danes. Their cultural and linguistic background may also contribute to the promotion of Danish exports to and imports from the
parts of the world they come from.
No matter how these things develop, as a liberal I am personally very concerned about the permanentand realoption of every
individual to choose his or her own way of life. I also believe it is the duty of the

< previous page page_8 next page >


< previous page page_9 next page >
Page 9
school system to work at the service of the nation, not vice versa. Nowadays we are discussing the schools' role in the formation
of the students' system of basic values. These include such notions as Human Rights, respect for and understanding of the
religious and moral beliefs of one's fellow human beings, and freedom of thought and word. It is crucial for me that these
concepts be taken seriously when it comes to the education of linguistic minorities, and I see this as an opportunity for both the
minority and the majority students. Perhaps one learns best about one's own national and cultural identity by having it compared
to somebody else's. This is also a road to global understanding and a feeling of global responsibility.
With this background it gives me great pleasure to be with so many distinguished colleagues and scholars from so many
different places. We have here representatives from countries which are very similar to Denmark, and with whom we share a lot.
We have also guests from far away who can tell us a lot. Furthermore, it is perhaps of particular importance that we have been
able to receive representatives from some of the countries which are the donors of the human resources that this nation has
received over the past 10-15 years. The necessary and fruitful exchange of knowledge, experience, thoughts, ideas, plans and
hopes should be encouraged as much as possible. The chance of this happening in a group of dedicated and distinguished
educators, such as you, is no doubt very high.
I wish you great success with your mutual presentation and exchange of material, and I would like to express the wish that this
conference become the starting point of a flow of international co-operation. With these words I declare the fifth Nordic
Conference on Bilingualism opened.

< previous page page_9 next page >


< previous page page_10 next page >
Page 10

< previous page page_10 next page >


< previous page page_11 next page >
Page 11

Bilingualism in Education and Bilingual Education:


The State of the Art in the United States
Anna Uhl Chamot
InterAmerica Research Associates,
Rosslyn, Virginia, U.S.A.
Abstract. This paper provides an overview of the current status and future directions of education for limited English
proficient (LEP) students in elementary and secondary schools in the United States.
First, the context of current educational activities is established through a brief description of the following topics:
estimates of numbers of LEP school-aged children;
legislative history of bilingual education;
estimates of qualified teachers for LEP students;
role of the Department of Education in the education of LEP students.
Following this, issues related to current U.S. government initiatives in the education of LEP students are discussed. The
effects of these initiatives on the following areas are described:
types of school programmes now eligible for federal funding;
network of services being provided to school districts serving LEP populations;
teacher training programmes,
recent research findings that are affecting the education of LEP students.
Finally, an assessment of probable future directions in the education of LEP students in the United States is discussed.
These directions include:
research on effectiveness of different methods for teaching English as a second language (ESL);
development of content-based ESL programmes and materials;
training of regular classroom teachers to meet the needs of LEP students;
efforts to improve the quality of programme evaluation;
research on second-language learning strategies.

Introduction
The United States of America has historically been a nation of immigrants, so that it is not surprising that languages other than
that of the early English

< previous page page_11 next page >


< previous page page_12 next page >
Page 12
settlers should have affected both individuals and institutions. The issue of bilingualism as it affects education is not new in the
United States, but at present the issue may be more visible than at other periods of history, because for many it has become a
political concern. As an educator, I would like to leave politics aside, insofar as it is possible, and discuss instead the educational
needs of our language minority school-aged population and describe what is being done at various levels to meet these needs.
The purpose of this paper is to provide a general overview of the education of limited English proficient (LEP) students in
American elementary and secondary schools as it exists today, to describe recent research in the field, and to indicate likely
future directions based on current and planned research.

Historical Background of Bilingual Education in the U.S.A.


At least since the beginning of this century, public education in the United States has been faced with the task of educating
immigrant children who do not speak English as their native language. In the early part of the century, immigrants were given
intelligence tests and many were proclaimed 'feeble-minded' because of their low scores on tests administered in English
(Cummins, 1984; Hakuta, 1986). Many of the children of these immigrants did not fare well in school; we remember the
brilliant exceptions who learned English quickly and went on to make substantial contributions to American society. But the
children who did not learn English easily dropped out of school and were absorbed into the labour force in factories or on farms.
In the early years of this century, many students did not finish secondary school, and the fact that among these dropouts there
was a disproportionate number of minority language students was not then recognised as a problem (Wong Fillmore, 1985).
Today the situation is different. Without at least a high school education, a person is nearly unemployable. Technology demands
a minimum level of education, and this minimum level is not obtainable without basic proficiency in English. Therefore, the
problems faced by schools today are quite different from those encountered 50 or 80 years ago. Today the schools are charged
with providing a basic education to the age of at least 16 in most states, and preferably to the point of high school graduation, at
the age of approximately 18. The largest single group of language minority students and the one for which we have the greatest
amount of school-related data, consists of Hispanic students. About half of Hispanic students drop out of school before
completing a high school degree (Duran, 1983), and in some places, such as New York City, the dropout rate is 80% (Bennett,
1986).
For the last 20 years, the United States government, through legislation and funding, has been seeking to meet the educational
needs of limited English proficient students, whether Hispanic or of more than 80 other

< previous page page_12 next page >


< previous page page_13 next page >
Page 13
ethnolinguistic groups (National Clearinghouse for Bilingual Education, 1985a). In order to understand the history of educational
movements in the United States, it is important to remember that, unlike many other countries, there is no central Ministry of
Education. Each state has its own education agency to guide educational matters, and a great deal of freedom in matters of
content and method is allocated to individual school districts within each state. Thus, the curriculum can vary not only in the 50
states, but also in individual school districts within a state. The Federal Government's role in education has always been one of
advice, research, provision of extra funds to school districts meeting certain criteria for educational needs, and of legislation to
overcome inequities due to race or national origin of students.
In 1967 the U.S. Congress amended the Elementary and Secondary Education Act by adding Title VII or the Bilingual
Education Act. The Bilingual Education Act is a demonstration programme which indicates that local school districts may apply
for funding to provide programmes to meet the needs of LEP students from minority language backgrounds, that is, students
from homes where a language other than English is dominant. The Federal Bilingual Education Act was a stimulus for a number
of states, such as California, Illinois, Massachusetts, Michigan and Texas, to provide their own bilingual education legislation
and programmes for their sizeable populations of LEP students.
A few years after the Federal legislation, litigation brought against the San Francisco school system was appealed to the level of
the Supreme Court. This case argued that the schools in San Francisco, which at the time did not offer any special programmes
for non-English speaking students, were not providing equal educational oppportunities for Chinese students who did not speak
English. What has come to be known as the Lau Decision states that the civil rights of LEP students must be protected by
offering them special programmes to meet their educational needs. Thus, in the late 60's and early 70's, important initiatives
affecting the education of LEP students came from both the legislative (Congress) and the Judicial (Supreme Court) branches of
the United States government. In each branch, these initiatives led to the establishment of bilingual programmes in many
schools.
The Bilingual Education Act was re-authorised by Congresss in 1974, and in this re-authorisation the law indicated that schools
may apply to receive financial assistance to conduct transitional bilingual education for their LEP students. The law made it
clear that bilingual education was to be transitional in nature, and defined the programme as follows:
The term 'program of bilingual education' means a program of instruction, designed for children of limited English-
speaking ability in elementary or secondary schools, in which, with respect to the years of study to which such program is
applicable(i) there is instruction given in, and study of, English and, to the extent necessary to allow a child to progress
effectively through the educational system, the native language of the children of limited Englishspeaking ability, and
such instruction is given with appreciation of the cultural

< previous page page_13 next page >


< previous page page_14 next page >
Page 14
heritage of such children, and with respect to elementary school instruction, such instruction shall, to the extent necessary,
be in all courses or subjects of study which will allow a child to progress effectively through the educational system . . .
(Bilingual Education Act, 1974: Sec. 703, (a) (4) (A)).
In other words, while school districts were free to offer non-bilingual types of programmes, they could get funding from the
government only by offering bilingual programmes. In 1975, the U.S. Office for Civil Rights published the 'Lau Remedies', or
guidelines that school districts should follow in order to comply with the Lau decision made by the Supreme Court in 1974. The
Lau Remedies (which were never made into a law) specified that Transitional Bilingual Education (TBE) had to be offered to
LEP students, and also stated that English as a second language (ESL) programmes alone could not be substituted for bilingual
education. Thus, school districts were encouraged to institute bilingual programmes for two reasons: (1) bilingual programmes
were the only programmes that were eligible for government funding; and (2) bilingual programmes were recommended by the
Office for Civil Rights to provide the equal education opportunities to LEP students required by law.
Thus began a period of rapid expansion of bilingual education in the United States. The rationale of providing comprehensible
instruction through a student's native language until the student acquired sufficient proficiency in English to use English as the
instructional medium made sense educationally: start with where a student is, then lead the student to where he or she needs to
be. A number of states at this time passed laws requiring schools to offer bilingual classes for LEP students when a certain
number of students was identified at a given grade level. Also at this time, a number of researchers began to write and talk
about the psychological and social constructs underlying bilingual education (Cummins, 1980, 1982; Lambert, 1980, 1981;
Tucker, 1980). The transferability of both knowledge and processes acquired in the first language (L1) to the second language
(L2) became the most important theoretical principle underlying bilingual education. While theoreticians were analysing
relationships between L1 and L2, the schools were dealing with the practical difficulties of providing something called
'transitional bilingual education' to many different types of students.
Adding to the practical implementation difficulties was the controversial nature of the bilingual education legislation. Many
people, whether educators, researchers, policy makers, or even parents of LEP children, were critical of bilingual education, both
for educational (it delays the acquisition of English) and for societal (it fosters political divisiveness) reasons (Wong Fillmore,
1985). When the proposed Lau Guidelines were discussed in public forums around the country, many of these objections to the
recommendations for bilingual education were raised (Wong Fillmore, 1985). In 1985, the new Secretary of Education, William
J. Bennett, instituted a

< previous page page_14 next page >


< previous page page_15 next page >
Page 15
number of changes in the Federal role in bilingual education. First, the Lau Remedies were withdrawn, and 'school districts were
informed that they have considerable discretion under the statute to determine the extent of native language instruction required
in a transitional bilingual education project' (Bennett, 1986:vi) Second, increased parental involvement was required, and third,
school districts were required to submit a plan for continuing special programmes for LEP students after government funding is
no longer available. Of these three objectives, only the first represented a significant shift in policy. The effect of allowing
school districts to decide how much native language instruction would be provided in transitional bilingual education
programmes was to encourage programmes which significantly diminished or eliminated the use of native language instruction.
The immediate result was an increase in the number of applications for Title VII funds for programmes in a category termed
'Special Alternative Instructional Programs', characterised by approaches which do not utilise a non-English language for
instructional purposes. However, funding from Congress for such programmes was limited to a maximum of 4% of all awards
for instructional programmes under Title VII. As a result, in 1985, of the 104 applications received for Special Alternative
Programs, only 35 were funded.

The Role of the Department of Education


In passing the Bilingual Education Act, the Congress also allocated funds to be used for instruction, teacher training, curriculum
development, research and evaluation of bilingual education. The Office of Bilingual Education and Minority Languages Affairs
(OBEMLA), in the U.S. Department of Education administers these funded programmes, which were initially designed to
provide assistance to school districts in dealing with the educational needs of LEP students. The following types of assistance
were made available at the time of the re-authorisation of the Bilingual Education Act in 1974 (Leibowitz, 1980):
Grants to school districts to assist in setting up bilingual programmes;
Grants to institutions of higher education to prepare teachers to work with LEP students;
Fellowships to graduate students for advanced study in the education of LEP students;
Resource centres to develop materials, provide inservice training of teachers, assist in asessment, and disseminate
information about bilingual education; and
Contracts for research activities to: identify the number of students requiring special educational services, find out how
many teachers have appropriate training to teach LEP students, study effective programme

< previous page page_15 next page >


< previous page page_16 next page >
Page 16
models, describe second language acquisition, identify effective methods of teaching English, operate a national clearinghouse
for information on bilingual education, study effective methods of teaching reading in a second language, describe effective
teacher training programmes and identify critical cultural characteristics of language minority populations.

Each of these areas has accomplished much with the aid of government support; each area has also encountered difficulties, and
none has discovered simple and unequivocal answers to meeting the educational needs of LEP students.
One of the most important endeavours at the beginning of the bilingual education movement was to document the actual number
of LEP students requiring special assistance. 'Limited English proficient' refers to individuals who have difficulty understanding
instruction in English, and, as defined by the Bilingual Education Act, applies to the following:
persons born outside the United States or persons whose native language is not English;
persons in whose environment a non-English language is dominant; and''
American Indians and Alaskan Natives in whose environment a non-English language has significantly affected their
proficiency in English.

The first estimate of the number of school-aged children with LEP was made by the Children's English and Services Study
(CESS) in 1978. This study estimated that 2.4 million school-aged children aged 5-14 were limited in English proficiency, and
that a total of 3.6 million children between the ages of 4 and 18 were limited in English (O'Malley, 1982). A second estimate of
the numbers of LEP school-aged children was made by the English Language Proficiency Survey (ELPS) in 1982. This survey
found that only 1.2 to 1.7 million school-aged children were limited in English. The differences between the two surveys are the
result of the establishment of different cut-off scores separating limited from proficient students in English, and the inclusion or
exclusion of students dominant in English but still limited in English proficiency. These two surveys reflect the difficulties of
ascertaining a child's language proficiency through interview data. Currently the Department of Education is using the lower
estimate of LEP students needing special educational services (Bennett, 1986).
Another important area of investigation has been the determination of the number of qualified teachers for LEP students. A
Teacher Language Skills Survey (TLSS) was conducted in 1976-77 and again in 1980-81. The earlier survey found that, of the
sample of elementary school teachers surveyed nationwide, only 31% had received some training in bilingual education or
English as a second language. The second TLSS found that in 1980-81 the

< previous page page_16 next page >


< previous page page_17 next page >
Page 17
number of such teachers had risen to 55% (O'Malley, 1983). In 1983-84, the National Longitudinal Evaluation (discussed in
more detail below) indicated that the percentage of elementary teachers with some training in bilingual education or ESL had
risen to 60%. The amount of training varied from fully qualified bilingual or ESL teachers with credentials satisfying state
certification requiremens (in states requiring such certification) to teachers who had had one or more university courses or
inservice training workshops related to teaching minority language students. That there is a scarcity of qualified bilingual
teachers in areas with large LEP populations is evidenced by recent, well-publicised shortages of qualified teachers in cities such
as Houston, Los Angeles, Boston, Seattle, San Francisco, Albuquerque, Phoenix, New York City and Washington, DC (Bennett,
1986).

Current Initiatives in Bilingual Education


The programmes currently administered by the Office of Bilingual Education and Minority Language Affairs (OBEMLA)
include instructional programmes, teacher training, a network of services for both school districts and individuals, and a number
of research activities.
Instructional programmes
Table 1 identifies the six different instructional programmes funded in 1985-86 under Title VII of the Bilingual Education Act,
the number of awards for each, the number of proposals received for each category, and the amount of funding awarded to each.
The greatest number of these programmes (538) is in the category of Transitional Bilingual Education, which consists of
structured English language instruction (ESL) and (to the extent necessary) instruction in children's native languages.
The next largest group receiving funding (37) are Academic Excellence Programs. This funding is for established programmes
(whether TBE, Special Alternative, or Developmental) which have already demonstrated that
Table 1 Bilingual Education Act programmes funded in 1985-86
Programme type Awards Funding ($)
Transitional Bilingual Education 538 77,342,400
Academic Excellence 37 7,750,000
Special Alternative 35 5,370,600
Special Populations 27 3,500,000
Family English Literacy 4 500,000
Developmental Bilingual Education 2 250,000
Totals 643 94,713,000
Source: Bennett, 1986.

< previous page page_17 next page >


< previous page page_18 next page >
Page 18
they are effective; the purpose is to disseminate information about these exemplary programmes so that they can serve as models
to others.
The next category is Special Alternative Programs, with a total of 35 projects receiving funding. These are not bilingual
programmes, but have special curricula designed to meet the educational needs of LEP students. These curricula at present
include extended ESL instruction for several hours a day (High Intensity Language Training), and immersion, sheltered English,
or other content-based programmes, in which instruction in all areas of the curriculum (not just language) is provided through
simplified English and contextualised activities (Chamot & Stewner-Manzanares, 1985).
The next category of instructional programmes includes 27 programmes for Special Populations. These programmes offer special
instruction for LEP students who are in pre-school, who are gifted and talented, or who are in special education settings. Four
Family English Literacy Programs are funded to meet the English and literacy needs of LEP adults and young people who are
not in school. In these programmes, preference is given to parents and other family members of children who are in other
programmes funded by Title VII. Finally, two programmes are funded for Developmental Bilingual Education. These
programmes are designed to develop competence in both English and another language, as well as provide for instruction in all
other areas of the curriculum.
In addition to the six categories of instructional programmes funded under Title VII, OBEMLA also administers two
programmes funded by other legislation. These are the Emergency Immigrant Education Program and the Transitional Program
for Refugee Children. Both of these programmes provide grants directly to states, the amount based on the number of eligible
students in the state concerned. The Emergency Immigrant Education Program provided funds to assist in the education of
422,549 students in 1985-86, distributed in the states of California, New York, Texas, Illinois, Florida and Massachusetts. Funds
from the Transitional Program for Refugee Children served the needs of 82,000 students in the same year, and were distributed
in the states of California, Florida, Texas, Virginia, Massachusetts and Illinois.
Teacher training programmes
Two types of grants are provided to assist in the preparation of teachers and other educational personnel to work with LEP
students. Grants are provided to institutions of higher education to train undergraduate and graduate students in bilingual
education, and also to provide fellowships for graduate students to pursue advanced study in areas related to the education of
LEP students. In 1985-86, grants were provided to 144 different university bilingual training projects, and 514 bilingual
education fellows were funded for advanced study.

< previous page page_18 next page >


< previous page page_19 next page >
Page 19
Title VII service network
A number of additional services are funded through Title VII to help states, school districts and individuals provide better
educational services to LEP students.
Multifunctional Resource Centres provide technical assistance, information, and inservice teacher training in 16 geographical
regions.
Two Evaluation Assistance Centres provide technical assistance in assessment and evaluation of LEP students and programmes
for states and local school districts.
Two instructional Materials Programmes are developing instructional materials in languages for which no such materials are
available from commercial publishers.
The State Educational Agency Programme provides grants to states to collect, analyse, and disseminate information on the LEP
population, including educational services provided, in each state.
The National Clearinghouse for Bilingual Education collects, analyses, and disseminates information about bilingual education
and related programmes.

Recent Research in Bilingual Education


In addition to assessment of national needs in bilingual education through surveys to estimate numbers of LEP students and
numbers of qualified teachers for such students, recent research in bilingual education and related areas has also investigated
effective classroom practices, described the range of educational services for LEP students, and reviewed programme
evaluations to ascertain the effectiveness of bilingual education. This section dicusses first the research on effective practices,
then studies describing educational services and, finally, current research on bilingual programme effectiveness.
Studies of effective classroom practices
The 'Descriptive Study of Significant Bilingual Instructional Features' (SBIF) was conducted from 1980 to 1983 by William
Tikunoff, with ancillary studies conducted by a consortium of eight agencies and institutions. The purpose of the main study was
to describe the classroom features of effective bilingual instruction in nine sites which included students whose ethnolinguistic
background was Puerto Rican, Cuban, Mexican, Navajo, Chinese, Vietnamese, Filipino and other cultures (National
Clearinghouse for Bilingual Education, 1985b).

< previous page page_19 next page >


< previous page page_20 next page >
Page 20
The main descriptive study applied the findings of previous research on effective schools and practices to bilingual education.
Fifty-eight bilingual teachers, who had been identified as effective by their peers, supervisors, parents and students, were
observed to discover how they delivered instruction to a total of 232 LEP students. The study found that five instructional
features were significant in providing successful instruction for LEP students. Features that characterised effective bilingual
teachers were:
use of active teaching behaviours, such as: giving clear directions, describing task objectives accurately, indicating expected
achievement, using a number of strategies to present information, pacing instruction, involving students actively, monitoring
student progress and providing appropriate feedback;
using both the native language and English to clarify and mediate instruction;
integration of English development into all subject areas of the curriculum, including those subjects taught in the L1;
building on the students' native culture(s) to make instruction more meaningful; and,
establishing objectives which were congruent with what the school expected students to achieve.

Most of the features identified paralleled previous research findings on the components of effective teaching. The two features
unique to bilingual education settings were the use of the native language to clarify instruction, and the use of the native culture
to make instruction meaningful to students.
A number of special studies related to the SBIF study took place at the same time, and with collaboration and exchange of
information between the researchers and institutions involved.
The study 'Learning English through Bilingual Instruction' was conducted by Lily Wong Fillmore for the purpose of
determining effective instructional practices in developing the English academic language skills of LEP students. This three-year
study took place in ten bilingual and seven English-only third and fifth grade classrooms in which LEP students had had two to
three years of exposure to English. An analysis of the instructional practices of 20 teachers, and the effects of these practices on
the learning of over 150 Hispanic sand Chinese students was conducted, and four major instructional factors were found to be
significant (National Clearinghouse for Bilingual Education, 1986a). These were:
high quality teaching, characterised by consistency in lesson organisation, comprehensibility of directions and explanations, use
of appropriate instructional materials, emphasis on higher level skills, and opportunities for oral activities,

< previous page page_20 next page >


< previous page page_21 next page >
Page 21
high quality instructional language, characterised by clarity, coherence, use of contextualisation, paraphrasing, responding to
student feedback, and discussion of grammar and vocabulary;
effective classroom management which focussed on academic learning more than on non-academic activities, and
provision of equal opportunities for all students to practise English during class activities.

This study found a number of differences between the Chinese and the Hispanic students in the classrooms studied. Chinese
students seemed to learn best in structured, fairly quiet classrooms, and profited most from interaction with their teachers.
Hispanic students profited most from interaction with their peers, and were more sensitive to the quality of teaching and of
instructional language than were Chinese students.
Effective classrooms were characterised by a balance of activities that were teacher-directed or individualised. In the bilingual
classrooms, students learned best when teachers kept the two languages separate and did not translate.
Another large-scale study was 'Teaching Reading to Bilingual Children', conducted by Betty Mace-Matluck and Wesley A.
Hoover from 1978 to 1984. The purposes of this study were to describe and document bilingual children's language abilities,
reading instruction, and the relationship of instructional programme type to children's achievement. Longitudinal data obtained
from 20 schools in a number of bilingual communities provided information about reading instruction for a sample of 250
Spanish-English bilingual students in kindergarten through third or fourth grade. Major findings addressed a number of
instructional and research issues (MaceMatluck, 1985):
English oral skills developed more than Spanish oral skills over time.
The oral language proficiency test was inferior to teachers' ratings in estimating students' abilities in both languages.
Most entering kindergarteners had sufficient language development to begin reading instruction.
28% of the sample students were in instructional programmes which taught them to read initially in Spanish.
Similar practices were observed for both Spanish and English reading instruction. These practices focussed on decoding skills,
especially in early grades, and limited time was devoted to higher order comprehension skills.
Children entering school with superior oral language skills, whether in Spanish or in English, showed more growth in reading
comprehension over time.

Mace-Matluck (1985) interprets these and other findings of this study to indicate that instructional practices can positively affect
the achievement of

< previous page page_21 next page >


< previous page page_22 next page >
Page 22
minority language students, and that improvement in reading instruction is needed for these students to become effective
readers.
Studies describing educational services
The 'National Longitudinal Evaluation of the Effectiveness of Services for Language Minority Limited English Proficient
Students' is a five-year study (1983-88), conducted by Malcom B. Young, which is being carried out in two phases. The
descriptive phase was completed in 1984, and the second phase, in which an evaluation of the effectiveness of different types of
educational services for LEP students is being conducted, is in progress.
The two main purposes of the descriptive phase of the study were to estimate the number of elementary school LEP students
receiving instructional services related to their language needs, and to describe the range and features of these services. A survey
of 520 elementary schools in 19 states was conducted. This consisted of mail questionnaires and telephone interviews of school
districts with small numbers of LEP students, and site visits to districts with large numbers of students receiving services. The
site visits provided information on about 4,000 teachers of grades 1 to 5, and about 1,600 LEP students in grades 1 and 3. The
findings of the descriptive phase of the National Longitudinal Evaluation study are based on information obtained from districts
with large numbers of LEP students (National Clearinghouse for Bilingual Education, 1986b). The major findings during the
1983-84 school year are listed below.
An estimated 882,000 LEP students were enrolled in U.S. elementary schools (in some cases, the count is of LEP students
receiving special services, rather than total number enrolled in school)
The definition of 'LEP student' varied considerably from school district to school district, and even between schools
Ninety-seven per cent of the large school districts surveyed were providing instructional services tailored to meet the needs of
LEP elementary school students
The instructional services provided varied widely in the amount of native language instruction offeredabout 14% of the schools
used a significant amount of instruction through the native language (generally Spanish), while most programmes offered little
or no native language instruction
Spanish was the predominant non-English language in 63% of the schools surveyed, and a Southeast Asian language
(Vietnamese, Cambodian, Laotian, Hmong) was predominant in 14% of the schools
Nearly 60% of first and third graders in the schools studied had been born in the United States
The major goal of all school districts surveyed was to develop students' English proficiency so that they could participate
successfully in all-

< previous page page_22 next page >


< previous page page_23 next page >
Page 23
English instruction, and nearly all districts also stated as a goal the development of academic skills necessary for school
achievement
Only 15% of the schools surveyed indicated that one of their goals was to develop or maintain proficiency in the students'
native language(s).

What emerges from the findings to date of the National Longitudinal Evaluation is that a number of educational services are
being provided for LEP elementary school students, and that in most cases these services are providing instruction mostly in
English, with only limited or infrequent use of the native language. However, these programmes are generally identified as
Transitional Bilingual Education programmes both in their titles and in their source of funding. The problem of accurate
descriptive labels has plagued a great deal of the research in bilingual education, both in the area of identification of LEP
students and in the evaluation of programme effectiveness.
In the descriptive research conducted on bilingual programmes and practices, only one study has focussed on the structured
English or ESL component of programmes for LEP students. The 'Review, Summary, and Synthesis of Literature on English as
a Second Language' (ESLIT), which I conducted in 1985, described the state of the art in ESL instructional approaches,
organisational patterns, instructional materials and second language learning theories in elementary and secondary schools. In
addition to reviewing and synthesising recent information on ESL in public schools, this study also addressed policy issues
related to ESL instruction. Information was gathered through a survey of recent literature on ESL in U.S. schools, supplemented
by interviews with a sample of researchers, ESL teacher trainers at major universities, and other ESL and bilingual education
professionals, including programme directors and teachers working with LEP students in school districts around the country.
The major findings of the ESLIT study were as follows (National Clearinghouse for Bilingual Education, 1985c):
Thirteen instructional approaches were reported to be used in schools, of which the most frequent are: audiolingual, eclectic (a
combination of approaches), Natural Approach, Total Physical Response, and communicative language teaching
Little evidence exists to support the effectiveness of any of the different ESL approaches for elementary and secondary school
students
Communicative approaches may be most appropriate for beginning level ESL students, especially young children, while
cognitive and content-based approaches may work better for developing academic language skills
Second language learning is facilitated in programmes which allocate a significant part of the school day to English language
development, which provide for interaction between LEP and English proficient students, and which provide native language
support

< previous page page_23 next page >


< previous page page_24 next page >
Page 24
Most available instructional materials are not designed to develop academic language skills in elementary and secondary
students
Few instructional approaches or materials identify their underlying language learning theories.

Although no major research on the effectiveness of different instructional approaches to teaching ESL in schools has been
completed since the ESLIT study, a number of changes are currently underway in the field. The feasibility of incorporating
content instruction with language development as a way to develop academic language skills is being explored by school
districts in many parts of the country. In addition, emerging research on the role of cognitive processes in second language
learning, discussed in the final section of this paper, has suggested instructional applications to the second and foreign language
classroom (Chamot et al., 1987; Chamot & O'Malley, 1986, 1987).
Studies of bilingual programme effectiveness
Evaluations of bilingual programmes have been criticised by both proponents and opponents of bilingual education for
methodological weaknesses. McLaughlin (1985:233) summarises these weaknesses as follows:
Definition of what is meant by bilingual education
Problem of brief duration of the study
Non-random assignment of subjects to treatment and control groups
Uncontrolled variable: socioeconomic status, intelligence, time on task, skill of teacher
Subject loss or attrition
Testing instruments not valid and/or reliable.
A review of major evaluation studies conducted in the last ten years illustrates how these weaknesses cloud the findings.
The first large-scale study to evaluate the effectiveness of bilingual education was carried out in 1977-78 by the American
Institutes for Research (Danoff et al., 1978). 11,500 students in Spanish/English Title VII-funded elementary programmes in 38
school districts around the country were compared to children not in bilingual programmes. One of the major purposes of this
study was to find out if the bilingual programmes led to improved English language skills for LEP students. The comparison
focussed on English (and Spanish) reading and on mathematics. Some of the students were observed for five months, and others
for a two-year period. The findings were that while students in bilingual programmes improved their Spanish reading skills, no
gains were found in either English or mathematics, compared to students in the control group.
Criticisms of the American Institutes for Research (AIR) study have pointed out that no degree of programme implementation
was considered,

< previous page page_24 next page >


< previous page page_25 next page >
Page 25
and comparison group students were not adequately matched in English language proficiency to those in the treatment groups
(O'Malley, 1978). In other words, there was no descriptive information about the programmes considered 'bilingual'amount of
time and type of instruction in each language, qualifications of teachers, appropriateness of instructional materials. A second
major criticism, which concerned English language proficiency, emerged when reviewers of the study noticed that the
percentage of the control group students who were already fluent in English far exceeded the percentage in the treatment group.
The fact that programme success was judged only on the basis of student achievement in English and mathematics has also been
criticised for narrowness of criteria (Hakuta, 1986). These two subjects, while of basic importance in the curriculum, should not,
many believe, be used as the sole yardstick of a programme's success. Paulston (1980), for example, states that a more
convincing measure would be comparison of dropout rates of minority language students in bilingual and in non-bilingual
programmes. Numerous other criticisms have been made of this study, but, until the release of the seond phase of the National
Longitudinal Evaluation, it stands as the only large-scale study that collected original data on the effectiveness of bilingual
education (Hakuta, 1986).
At least three investigations have analysed and synthesised the findings of individual programme evaluations in order to
aggregate findings on effectiveness of bilingual programmes. The first of these was the Baker & de Kanter (1981) study, which
reviewed over 300 individual evaluations and selected from them 28 which met criteria established by the researchers for the
technical quality of the evaluation. These 28 evaluations were analysed to discover whether children in bilingual programmes
developed English language skills better than children in non-bilingual programmes. Their findings were that while special
programmes for LEP students can improve their achievement, 'the case for the effectiveness of transitional bilingual education is
so weak that exclusive reliance on this instruction method is clearly not justified' (p.1). The authors recommended immersion
programmes, defined by them as 'structured curriculums in English for both language and non-language subject areas' (p.2) as an
alternative to bilingual programmes for LEP students.
A number of criticisms of the Baker-de Kanter report have been made. The criteria used to select the 28 evaluation studies
analysed and the ways in which the criteria were applied have both been attacked by the American Psychological Association
(1982). The recommendation of 'immersion' as an alternative was based on a single kindergarten programme in its first year,
which should not have been considered an immersion programme according to the Baker-de Kanter definition because children
received reading instruction in Spanish (McLaughlin, 1985). In spite of these and other criticisms, however, the Baker-de Kanter
report has been influential in guiding the policy of the Department of Education for the education of LEP students as evidenced
in three related events: (1) a large-scale study has been funded

< previous page page_25 next page >


< previous page page_26 next page >
Page 26
to compare immersion to bilingual programmes (2) 35 Special Alternative Programs which provide immersion or other types of
ESL instruction are currently funded; and (3) the Department is seeking Congressional approval to lift the current cap of 4% of
total funding for such alternative programmes.
The second synthesis study was one in which Ann Willig (1985) conducted a meta-analysis of the same studies analysed by
Baker and de Kanter, but eliminated studies which had not been conducted in the United States or were not part of a regular
school programme. The 23 studies remaining were subjected to a statistical synthesis, using meta-analysis methods. The
variables which accounted for most of the variance in effect sizes across studies were identified. When considered together,
analyses of these variables indicated that 'there were overall significant, positive effects for bilingual education programmes both
for tests administered in English and tests administered in Spanish' (Willig, 1985, p.277). An additional finding was that the
quality of evaluation design had important effects on the results. For example, some studies included in the comparison group
students who had successfully completed bilingual programmes and were functioning successfully in the mainstream. In other
studies, the student attrition rate in bilingual programmes between pre- and post-tests lowered post-test scores because better
students left the bilingual programmes and were replaced by students who were more in need of special assistance. Studies in
which the research methodology was of better quality showed the most positive effects of bilingual education. Willig (1985)
criticised the general inadequacy of research on effectiveness of bilingual education. She called for comparisons of different
types of programmes for LEP students, since comparisons with randomly assigned controls are not legally possible due to
requirements that LEP students be provided with special instructional services.
The most recent major study to analyse the effectiveness of bilingual education was conducted by the General Accounting
Office (GAO), the U.S. government agency charged with providing information to Congress on the effectiveness of government
programmes and activities. GAO's mandate includes the 'review of the policies and practices of federal agencies administering
education programmes' (General Accounting Office, 1987, p.26). The GAO study was conducted in response to a request from
the Congressional committee which oversees educational issues. As mentioned above, the Department of Education is currently
seeking Congressional approval that the requirement of native language teaching be dropped from the Bilingual Education Act
so that more than the current 4% of funding allocated to Special Alternative Programs can be used to fund additional English-
only programmes. To support this position, the Department of Education has cited research such as the Baker-de Kanter report
and claimed that the research on bilingual education's effectiveness is too inconclusive to warrant major reliance on this
approach to educating LEP students. Statements by the Secretary of Education and other officials in the Department have
received a great deal of attention from both educators and the general

< previous page page_26 next page >


< previous page page_27 next page >
Page 27
population. In response to these statements, which are seen by many as indicating a policy shift, the GAO was asked by
Congress to investigate the validity of the claims being made by officials in the Department of Education. The GAO asked ten
nationally recognised experts in bilingual education, general education and educational research to examine these official
statements and judge their validity, based on ten reviews of effectiveness of different approaches to the education of minority
language children. The reviews included the Baker-de Kanter report, a response to this report (Yates & Ortiz, 1982), two
reviews on immersion methods (Gersten & Woodward, 1985; Hernandez-Chavez, 1984), and six general reviews, including the
Willig meta-analysis summarised above.
The results of the GAO study, released in April 1987, state that
First, only 2 of the 10 experts agree with the department that there is insufficient evidence to support the law's requirement
of the use of native language to the extent necessary to reach the objective of learning English. Second, 7 of the 10 believe
that the department is incorrect in characterizing the evidence as showing the promise of teaching methods that do not use
the native language . . . Few agree with the department's general interpretation that evidence in this field is too ambiguous
to permit conclusions (General Accounting Office, 1987:3).
In general, while acknowledging the weaknesses in research and evaluation in bilingual education, the majority of the experts
indicated that there is enough evidence about bilingual education research to support the legal requirement of native language
instruction in federally funded programmes for LEP students.
The GAO study was criticised methodologically by the Department of Education in two main areas. First, the reliance on the
opinions of a limited number of experts may have introduced subjectivity and even bias into the findings. GAO acknowledges
this and indicates that the experts did represent different views on the issues, that all are highly qualified and respected in their
respective areas, and that more than half of them had been cited by the Department of Education in support of its policies. A
second area in which the GAO study was criticised by the Department is its use of reviews of studies on the effectiveness of
bilingual and other special language programmes, rather than referring to original sources. The reason for the use of reviews was
the limitation in time allowed for this study by the congressional committee that requested it. The ten reviews used were
selected after reviewing 52 analyses of multiple studies which had been identified from the abstracts of a total of 929 relevant
documents. Outside reviewers further narrowed the list, and additional criteria, such as recency, impartiality, breadth of
coverage, diversity of teaching approaches and number of learning outcomes, were used to select the ten reviews sent to the
panel of experts.
In summary, the research on effectiveness of bilingual education in the United States leaves a great deal to be desired. While
methodological weak-

< previous page page_27 next page >


< previous page page_28 next page >
Page 28
nesses have been cited by both supporters and critics, only recently has any assistance been made available to school districts
conducting evaluations of their bilingual programmes. The technical assistance which is currently being provided by the two
Evaluation Assistance Centres to advise school districts on designing and conducting evaluations should ensure a larger number
of methodologically sound programme evaluations in the future.

New Directions
This final section discusses current studies in bilingual education and second language learning which have the potential for
advancing our understanding of how best to meet the needs of LEP students in U.S. schools. These studies have been funded
from a variety of sources, ranging from those undertaken by private individuals to those funded through OBEMLA and other
federal entities.
In the first category, privately funded studies, a recently completed study by Virginia Collier (1987) has, I believe, important
instructional implications for LEP students in the United States. This study investigated the relationship between age at arrival,
prior education, and acquisition of academic skills of immigrant students in a large suburban school district. The purpose of the
study was to determine the time required for these students to develop academic language skills sufficient for satisfactory school
achievement. Collier analysed the standardised achievement scores of 1,548 minority language students who graduated from an
ESL programme between 1977 and 1986, and who, except for 5-7 year olds, had been at grade level in their native countries
upon arrival in the United States. The ESL programme was characterised as consisting of special instruction in English language
arts for part of the day, including 'some' assistance in content areas, and mainstream instruction for the remainder. Most students
left the ESL programme within two to three years. Collier reported that students who had entered the ESL programme between
the ages of 8 and 11 had taken two to five years (1-3 years in ESL followed by 1-3 years in mainstream instruction) on the
average to approximate the 50th percentile on nationally standardised tests of reading, language arts, science and social studies.
In contrast, students who had entered the ESL programme at ages 5 to 7 without prior schooling were significantly behind those
who had started their initial education in their native countries. Finally, students who entered the ESL programme at ages 12 and
above were only at about the 40th percentile on most tests after four years of instruction in English (ESL and mainstream
instruction). Collier argued that students who entered at ages 12 and above encountered the heavy cognitive academic language
demands of the secondary school and consequently were liable to take longer to show progress on grade-appropriate
standardised tests in English.
Collier's findings can be interpreted as supporting previous research by

< previous page page_28 next page >


< previous page page_29 next page >
Page 29
Cummins (1982, 1984) on the time lag between the development of social interactive language skills, as promoted by many ESL
programmes, and the full development of academic language skills. Collier indicates that her findings also argue in favour of
native language instruction while students are developing English skills.
If we turn to federally funded research, a number of current research activities sponsored by the Office of Bilingual Education
and Minority Languages Affairs (OBEMLA) seek to inform policy in bilingual education in three areas: (1) description of
student characteristics and educational outcomes; (2) evaluation of bilingual teacher training programmes; and (3) comparison of
effectiveness of current instructional programmes for LEP students.
Current projects which are analysing student characteristics and educational outcomes are using databases developed by the
Center for Educational Statistics in the Department of Education. OBEMLA provides funding to the Center for increasing the
sample size of Hispanic and Asian-American students in selected studies in order to estimate the size of the LEP student
population, their educational needs, and outcomes such as high school graduation and college enrolment. These projects are
focussing on secondary level LEP students. OBEMLA also plans to fund small research projects to investigate whether gifted
students and those with learning disabilities who are LEP are receiving adequate educational services.
The second area of OBEMLA research initiatives, the quality and effectiveness of teachers and teacher education programmes,
is intended to determine the current supply and demand of teachers for bilingual programmes (the last national survey of
teachers was conducted by O'Malley in 1980-1981 (National Clearinghouse for Bilingual Education, 1983)), and to investigate
the standards, curricula, attrition and graduation rates of bilingual teacher training programmes funded through Title VII.
Currently, OBEMLA has an agreement with the Center for Educational Statistics to provide supplemental information on
bilingual education teacher graduates within national surveys which are investigating teacher supply and demand and
characteristics of recent college graduates. OBEMLA has, in addition, funded a study to evaluate teacher training programmes.
This study will analyse the curricula and textbooks used in bilingual teacher training programmes, the methodologies taught
(whether bilingual education or ESL), and programme characteristics such as selection, requirements, graduation rates,
backgrounds of students and faculty, and employment of graduates.
In the third research area, effectiveness of instructional programmes, OBEMLA is involved in three major studies. The first is
the 'National Longitudinal Evaluation of the Effectiveness of Services for Language Minority Students', which has already
described the services provided to elementary school LEP students, and is now studying their degree of effectiveness.
A second longitudinal study is comparing the relative effectiveness of three instructional approaches over a four year period. The
approaches

< previous page page_29 next page >


< previous page page_30 next page >
Page 30
under study are: (1) immersion strategy programmes (which have been equated with 'sheltered' programmes, in which content
subjects are taught through simplified English); (2) early-exit or short term (two to three years) transitional bilingual education
programmes; and (3) late exit or long-term (five to six years) transitional bilingual education programmes (Baker & Ramirez,
1987). At the end of the first year of the study, students in the long term bilingual programmes had made more progress in both
Spanish and English than had students in the other two groups (Hakuta & Gould, 1987). One of the findings of the second year
of the study was that the amount of English used by students in school was related to the type of instructional programme. Not
surprisingly, English is used more frequently by both students and teachers in immersion programmes, and least frequently in
the late-exit bilingual programmes (Baker & Ramirez, 1987).
The third OBEMLA project currently underway is a study on the selection and exit procedures for students in bilingual
programmes. This study is investigating the effectiveness of instruments and procedures used to identify children in need of
bilingual education, and to determine when children are able to leave a bilingual programme in order to participate successfully
in mainstream classes. Another current study in this area concerns a survey of parents' attitudes and opinions about bilingual
education.
Planned OBEMLA studies in this area will investigate innovative instructional programmes for LEP students. Research to be
sponsored includes recent advances in cognitive psychology, linguistics, educational anthropology and foreign language
learning. Many of these studies will require the active collaboration of school districts in the research efforts.
Additional ongoing research is being conducted by the Center for Language Education and Research (CLEAR), which was
established in 1985 by the Office for Educational Research and Improvement of the Department of Education to conduct
research on education for LEP students and on foreign language education. CLEAR is currently conducting research projects in
these areas:
the development of academic skills in reading, writing and mathematics for language minority students;
the identification of cognitive and problem solving strategies for academic tasks;
the function of metalinguistic abilities in language acquisition, including transfer of knowledge;
the description of foreign language programmes and assessment of language proficiency,
the attrition of language skills, whether of first or subsequent languages; and
the features of interlocking foreign language programmes and programmes for LEP students.

< previous page page_30 next page >


< previous page page_31 next page >
Page 31
The results of these research initiatives are expected to provide valuable information on the educational needs of both minority
and majority language students in the United States. The studies which are focussing on language and academic skills learning
processes are undertaking much needed basic research on how second language learners acquire and utilise new linguistic and
conceptual information. The studies which are investigating the relationships between bilingual and foreign language
programmes will provide information about similarities and differences in these two types of language programmes, and ways in
which they can support each other.
A number of other government-sponsored studies are also investigating language learning processes in both minority and
majority language students. These studies provide additional information to the studies being conducted by CLEAR by focussing
on the learning strategies used by students acquiring a second or foreign language in a classroom context. Three of these current
studies have investigated the learning strategies of ESL students, and three are conducting parallel investigations of majority
language students learning a foreign language. These six learning strategies studies are discussed briefly as a conclusion to this
section on new directions in research.
The 'ESL Learning Strategies Studies', conducted by J. Michael O'Malley from 1982-86, investigated the learning strategies used
by secondary school ESL students for various language tasks (O'Malley et al., 1985a; O'Malley et al., 1985b; O'Malley, Chamot
& Kupper, 1986). Learning strategies are special techniques that students use consciously to comprehend, store and retrieve new
information. In the studies summarised below, learning strategies of LEP students were investigated in the context of academic
language tasks, rather than social interactive exchanges. This focus was selected in order to explore potential applications of
learning strategies to the academic needs of LEP students.
In the Descriptive Study, (O'Malley et al., 1985a) 70 high school ESL students (predominantly Hispanic), classified by their
teachers as good language learners, were interviewed to discover the range and frequency of learning strategies they used for
various language tasks in and outside the classroom. Nineteen group interviews were conducted, and in these 26 different
strategies were described, and 638 instances of strategy use were recorded. Although teachers indicated that generally they were
not aware of the strategies their students used, students themselves were able to describe their own strategies in detail. Discrete
language tasks, such as vocabulary learning and pronunciation, accounted for most strategy use, and students used a greater
proportion of cognitive strategies than they did metacognitive strategies. The importance of this study was in the identification
of a wide range of strategies used by ESL students and in the discovery that these students were consciously aware of their own
mental processes while they were learning English.
In the Learning Strategies Training Study, (O'Malley et al., 1985b) 75 high school ESL students (Hispanic, Asian and other
ethnolinguistic groups), at

< previous page page_31 next page >


< previous page page_32 next page >
Page 32
the intermediate level of English proficiency and with mixed language learning abilities, were trained for two weeks to use
learning strategies on language tasks. Students were randomly assigned to experimental or control groups, and pre- and post-
tests were administered. The most striking result was that experimental (strategy training) groups significantly outperformed the
control group on the oral presentation task, which involved the use of academic language skills to prepare and present a brief
oral report. The experimental groups also did better on most listening tasks, but their success on these tasks varied with
difficulty level and amount of training cues. On the vocabulary task, Hispanic experimental subjects outperformed controls, but
Asian controls outperformed the Asian experimental group. A major finding in this study was that LEP students could be taught
to use learning strategies for academic tasks, and that this instruction could improve their performance.
The ESL Listening Comprehension Study (O'Malley et al., 1986) used think-aloud interviews with 11 ESL high school students
from Central America who had been rated as either effective or ineffective listeners by their teachers. Individual students were
given a listening task and had to report on the mental processes they consciously used to comprehend and remember the
information presented. On interviews in which comparisons were made between two groups, effective students had an average
of 74 strategy uses per interview, but ineffective students averaged only 44 strategies per interview. The three strategies used
significantly more often by students identified by their teachers as effective listeners were: self-monitoring for comprehension,
elaboration of meaning by relating information to prior knowledge, and inferencing of meaning through contextual clues. An
important part of this study was the development of a research model to examine the role of learning strategies and cognition in
second language acquisition (O'Malley, Chamot & Walker, 1987).
As a follow-up and comparison to the ESL learning strategies studies, Dr O'Malley and his co-workers designed and initiated a
three-year study on learning strategies in foreign language education. This project, which I am currently directing, includes three
inter-related studies (Chamot et al., 1987) that are investigating the learning strategies of secondary school students of Spanish
and of university level students of Russian.
In the Descriptive Study, (Chamot et al., 1987) questionnaires were administered to and interviews were conducted with 67
Spanish students and 34 Russian students at different levels of language study. This study paralleled the 'ESL Descriptive Study'
conducted previously. The major differences were that both ineffective and effective students participated, and that students at
three different levels of proficiency were interviewed. The range of strategies reported matched the range reported by ESL
students, and, in common with the ESL students, foreign language students reported using substantially more cognitive
strategies than metacognitive ones, though students at more advanced levels used a slightly larger proportion of metacognitive
strategies. An unexpected finding of this study was that students ident-

< previous page page_32 next page >


< previous page page_33 next page >
Page 33
ified as less effective language learners by their teachers nevertheless reported using a variety of learning strategies. The major
difference in strategy use between these ineffective students and students identified as effective language learners was that
effective students reported greater range and greater frequency of strategy use. An important finding of this study was that
majority language students use the same types of strategies when learning a foreign language as minority language students do
when learning English as a second language.
The Longitudinal Study (Chamot et al., 1987) is following 40 Spanish students and 13 Russian students at different proficiency
levels for four semesters. Data are collected through think-aloud protocols similar to the ones used in the ESL Listening
Comprehension Study, except that students work on tasks involving reading, writing and grammar, as well as listening
comprehension. This methodology is revealing insights into relationships between strategies and qualitative differences in the
ways in which effective and ineffective learners approach a language task. Preliminary findings indicate that more effective
learners use more metacognitive strategies and have a wider repertoire of strategies. In addition, effective learners are adept at
combining different strategies on a specific task. They monitor their own comprehension more and make more effective use of
prior knowledge than do less effective learners.
In the Course Development Study, (Chamot et al., 1987) initiated in April 1987, teachers are showing students how to use
learning strategies for some of the regular language activities in their classroom. This study is expected to provide additional
information on instructional applications of learning strategies.
In summary, investigations of learning strategies in second language education lead us to conclude that strategic learners are
more effective learners, that learning strategies can be taught, and that learning strategy instruction may be particularly valuable
in preparing limited English minority language students to participate successfully in the academic curriculum. Later in this
journal issue you will have the opportunity of reading Dr O'Malley's description of an instructional approach which we have
developed which uses learning strategy instruction to develop cognitive academic language skills in LEP students.

Conclusions
This paper has described the major features of bilingual education as it currently exists in the United States, with particular
focus on the role of the Federal Government in policy and research. Bilingual education and bilingualism in education are
perhaps beginning to be viewed within a larger framework of language education, including both English and non-English
languages. If this direction continues, bilingual education may one day be seen not as a compensatory programme for minority
language children, but as an enrichment programme for majority language as well as minority language students.

< previous page page_33 next page >


< previous page page_34 next page >
Page 34
References
American Psychological Association (1982) Review of the Baker-de Kanter report for the Congressional Hispanic Caucus.
Washington, DC.
Baker, K.A. and de Kanter, A.A. (1981) Effectiveness of Bilingual Education: A Review of the Literature. Washington, DC:
Office of Planning, Budget, and Evaluation, National Institute of Education.
Baker, K.A. and Ramirez, D. (1987). Becoming a more frequent speaker of a second language: Some effects of program model
on LEP students and their teachers. Paper presented at American Educational Research Association Annual Meetings, Washing,
DC, 23 April, 1987.
Bennett, W.J. (1986). The Condition of Bilingual Education in the Nation, 1986: A report from the Secretary of Education to the
President and the Congress. Washington, DC: Department of Education.
Bilingual Education Act (1974). P.L. 93-380, (21 Aug. 1974), 88 Stat. 503.
Chamot, A.U. and O'Malley, J.M. (1986) A Cognitive Academic Language Learning Approach: A Content-based ESL
Curriculum. Wheaton, MD: National Clearinghouse for Bilingual Education.
(1987) The cognitive academic language learning approach: A bridge to the mainstream. TESOL Quarterly, 21 No. 2., 227-49.
Chamot, A.U., O'Malley, J.M., Kupper, L. and Impink-Hernandez, M.V. (1987) A Study of Learning Strategies in Foreign
Language Instruction: First Year Report. Rosslyn, VA: Interstate Research Associates.
Chamot, A.U. and Stewner-Manzanares, G. (1985) A Review, Summary, and Synthesis of Literature on English as a Second
Language: Final Report. Rosslyn, VA: InterAmerica Research Associates.
Collier, V.P. (1987) Age and rate of acquisition of second language for academic purposes. TESOL Quarterly, 21, No. 4, 617-
41.
Cummins, J. (1980) The cross-lingual dimensions of language proficiency: Implications for bilingual education and the optimal
age issue. TESOL Quarterly, 14, No. 2, 175-87.
(1982) Tests, Achievement, and Bilingual Students: FOCUS 9. Wheaton, MD: National Clearinghouse for Bilingual Education.
(1984) Bilingualism and Special Education: Issues in Assessment and Pedagogy. Clevedon, England: Multilingual Matters.
Danoff, M.N. (1978) Evaluation of the Impact of ESEA Title VII Spanish/English Bilingual Education Programs: Overview of
Study and Findings. Palo Alto, CA: American Institutes for Research.
Duran, R. (1983) Hispanics' Education and Background: Predictors of College Achievement. New York: College Entrance
Examination Board.
General Accounting Office. (1987) Bilingual Education: A New Look at the Research Evidence. Washington, DC: U.S. General
Accounting Office.
Gersten, R. and Woodward, J. (1985) A case for structured immersion. Educational Leadership, 43, No. 1, 75-9.
Hakuta, K. (1986) Mirror of Language: The Debate on Bilingualism. New York: Basic Books.
Hakuta, K. and Gould, L.J. (1987). Synthesis of research on bilingual education. Educational Leadership, 44, March 38-45.
Hernandez-Chavez, E. (1984) The inadequacy of English immersion education as an educational approach for language minority
students in the United States. In Studies in Immersion Education. Sacramento, CA: California State Department of Education.
Lambert. W.E. (1980) The Two Faces of Bilingual Education: FOCUS 3. Wheaton, MD: National Clearinghouse for Bilingual
Education.
(1981) The Social Psychology of Language: A Perspective for the 1980s: FOCUS 5. Wheaton MD: National Clearinghouse for
Bilingual Education.
Leibowitz, A.H. (1980) The Bilingual Education Act: A Legislative Analysis. Wheaton, MD: National Clearinghouse for
Bilingual Education.

< previous page page_34 next page >


< previous page page_35 next page >
Page 35
Mace-Matluck, B.J. (1985) An investigation of the effects of the interaction of student characteristics and instructional practices
on the language and reading achievement of a sample of bilingual children in Texas: The SEDL bilingual reading study. Paper
presented at First Annual Bilingual Research Symposium, Pan American University, McAllen, Texas, 6-7 June, 1985.
McLaughlin, B. (1985). Second-Language Acquisition in Childhood, Vol. 2: School-Age children (second edition) Hillsdale, NJ:
Lawrence Erlbaum.
National Clearinghouse for Bilingual Education (1985a) Bilingual Education Information Packet. Wheaton, MD: National
Clearinghouse for Bilingual Education.
(1985b) Significant bilingual instructional features study. Factsheet. Wheaton, MD: National Clearinghouse for Bilingual
Education.
(1985c) A review, summary, and synthesis of literature on English as a second language. Factsheet. Wheaton, MD: National
Clearinghouse for Bilingual Education.
(1986a) Learning English through bilingual instruction. Factsheet. Wheaton, MD: National Clearinghouse for Bilingual
Education.
(1986b) The national longitudinal evaluation of the effectiveness of services for language minority limited-English-proficient
students. Factsheet. Wheaton, MD: National Clearinghouse for Bilingual Education.
O'Malley, J.M. (1978) Review of the evaluation of the impact of ESEA Title VII Spanish-English bilingual education programs.
Bilingual Resources, 1, 6-10.
(1982) Children's English and Services Study: Educational and Needs Assessment for Language Minority Children with Limited
English Proficiency in the United States. Wheaton, MD: National Clearinghouse for Bilingual Education.
(1983) The need for teachers in bilingual programs: Findings from the 1980-81 Teachers Language Skills Survey. FORUM, VI,
No. 3, 2-4.
O'Malley, J.M., Chamot, A.U., Stewner-Manzanares, G., Kupper, L., and Russo, R.P. (1985a) Learning strategies used by
beginning and intermediate ESL students. Language Learning, 35, No. 1, 21-46.
O'Malley, J.M., Chamot, A.U., Stewner-Manzanares, G., Russo, R.P., & Kupper, L.(1985b). Learning strategy applications with
students of English as a second language. TESOL Quarterly, 19, No. 3, 557-84.
O'Malley, J.M., Chamot, A.U., and Kupper, L. (1986) The Role of Learning Strategies and Cognition in Second Language
Acquisition: A Study of Strategies for Listening Comprehension Used by Students of English as a Second Language. Rosslyn,
VA: InterAmerica Research Associates.
O'Malley, J.M., Chamot. A.U., and Walker, C. (1987) Some applications of cognitive theory to second language acquisition.
Studies in Second Language Acquisition, 9, No. 3
Paulston, C.B. (1980) Bilingual Education: Theories and Issues. Rowley, MA: Newbury House.
Tucker, G.R. (1980) Implications for U.S. Bilingual Education: Evidence from Canadian research: FOCUS 2. Wheaton, MD:
National Clearinghouse for Bilingual Education.
Willig, A. (1985) Meta-analysis of studies on bilingual education. Review of Educational Research, 55, No. 3, 269-317.
Wong Fillmore, L. (with Concepcion Valadez) (1985) Teaching bilingual learners. In M.C. Wittrock (ed.) Handbook of
Research on Teaching Third edition. New York: Macmillan.
Yates, J.R., and Ortiz, A.A. (1983) Baker and de Kanter review: Inappropriate conclusions of the efficacy of bilingual education.
NABE Journal, 7, No. 3, 75-84.

< previous page page_35 next page >


< previous page page_36 next page >
Page 36

< previous page page_36 next page >


< previous page page_37 next page >
Page 37

Bilingual Education and Bilingualism in Education:


A Comment
Gabriele Kasper
Dept of Linguistics, University of Aarhus
8000 Aarhus, Denmark.
My comments will address the last section of Dr Chamot's paper: the research on learning strategies.
Since the early 1970's, second language acquisition researchers have been interested in the operations by which second language
(L2) learners increase their knowledge of L2 and make such knowledge better accessible. However, only fairly recently have
attempts been made to identify learning strategies by using learners directly as informants, rather than inferring their mental
operations from performance data. Such research, using 'introspective methods', or 'verbal report data', has yielded valuable
information that has not only enhanced our understanding of the cognitive and interactional operations that bring about learning
but also by indicating how learning strategies could be incorporated into second or foreign language teaching. The educational
orientation towards the process dimension of learners' communicative knowledge manifests itself in the recent emphasis on
listening and reading strategies in many foreign language programmes. Yet the finding that students' learning strategies can be
improved by instruction has much more fundamental educational implications. It gives empirical support to the proposal to
include the development of strategic competence among the goals for second and foreign language teaching. By strategic
competence I understand both the operations learners use to solve problems in receptive and productive L2 use, and learning
strategies, i.e. operations used to increase L2 knowledge and its accessibility. This notion of strategic competence as part of
learners' procedural knowledge is much more comprehensive than Canale & Swain's (1980) original concept (cf. Færch &
Kasper, 1985).
Strategic competence in this broad sense, including metacommunicative and metacognitive awareness about communication and
learning processes, can appropriately serve to implement general educational concepts such as learner autonomy, self-directed
and individualized learningit can be viewed as a specification of the principle non scolae sed vitae discimus in the context of
bilingual education. Strategic competence enables learners to

< previous page page_37 next page >


< previous page page_38 next page >
Page 38
continue L2 learning outside the classroom, and to master unforeseen communicative situations. From this perspective, projects
on learning strategies, such as outlined by Dr Chamot, contribute importantly to bilingual education.
Obviously, no research programme is flawless, and I would like to comment on what appears to me problematic about the
reported investigations. These are four aspects:
the restricted scope of learning strategies envisaged in the studies;
the lack of theory-relatedness;
the employed research methods; and
the cultural universality/specificity of learning strategies

1
The Restricted Scope of Learning Strategies
The quoted research examines learning strategies only in the context of academic language tasks, explicitly excluding social-
interactive types of language use. The decision thus to restrict the scope of learning strategies is based on the distinction
between basic interpersonal communicative skills (BICS) and cognitive/academic language proficiency (CALP), as proposed by
Cummins in a variety of papers (cf. e.g. 1979). As many linguists and educationalists have pointed out (e.g. Martin-Jones &
Romaine, 1986), compartmentalising human activity into 'communicative' on the one hand and 'cognitive' on the other is totally
unwarranted, and I need not recapitulate their arguments for this audience. Let me just point out that I do not think we should
use an opposition such as 'academic' versus 'social-interactional' language tasks as long as we are not sure what exactly we mean
by it. Now if 'academic language tasks' are meant to include all the language tasks a student should be able to cope with to get
through school successfully, social-interactional knowledge and skills are sure to be centrally involved in such tasks. Therefore,
even if one wishes to restrict the scope of learning strategies to academic achievement, social-interactional aspects play a
fundamental role, and therefore learning strategies directed towards such aspects will have to be part and parcel of learning
strategy research and training.
Surviving in foreign second language teaching at school is certainly a fair enough end in itself, and restricting the scope of
learning strategies to be subservient to this goal, perfectly legitimate. Yet I'd prefer educational programmes to include learning
strategies that can also be successfully applied outside the school context, in a variety of situations relevant to the individual
student, helping him her develop a variable, flexible, dynamic competence to meet differential cognitive and communicative
demands.

< previous page page_38 next page >


< previous page page_39 next page >
Page 39
2
Lack of a Theory for L2 Learning
A serious weakness of the learning strategies studies under discussion, and which they share with previous investigations such as
the OISE Good Language Learner Project (Naiman, et al., 1978) or Rubin's work on learning strategies (e.g. 1981), is that they
lack reference to an explicit model of L2 learning. Due to the absence of such a model no criteria can be established that allow
us to unambiguously identify a behaviour as a learning strategy, or to evaluate the efficiency of different strategies. I do not
think it is scientifically sound to accept as learning strategies whatever behaviour learners label as such.
Let me illustrate the theory-dependency of learning strategies by a few examples from the descriptive ESL study (O'Malley et
al., 1985). Table 1 indicates the status of some of the learning strategies listed by O'Malley et al. (1985) according to four
different models of L2 learning: Universal Grammar (see Cook, 1985 for an introduction to its use as a model for L2
acquisition), Krashen's Monitor Model (e.g. Krashen, 1981), the Hypothesis Formation and Testing Model (e.g. in the version of
Færch, Haastrup & Phillipson, 1984) and the Interactional Approach (e.g. Hatch, 1984 for an overview). The table should be
self-explanatory, but let us take one example as an illustration. Transfer from the native or another foreign language is fully
compatible with a Universal Grammar model of L2 learning, where
Table 1 Some learning strategies from O'Malley et al. (1985) and their
status in different models of L2 learning
Learning strategy Universal Monitor Hypothesis Interaction
Grammar Model Testing Approach
Delayed production ? ? No No
Directed physical No No No Yes
response
Repetition No No No Yes
Deduction Yes No Yes No
Imagery No No No No
Transfer Yes No Yes Yes
Inferencing ? Yes Yes Yes
Requests for ? Yes Yes Yes
clarification
Auditory No No No No
representation

< previous page page_39 next page >


< previous page page_40 next page >
Page 40
L1 knowledge is taken to interact with universal principles in the learners' construction of an L2 grammar (e.g. Gass, 1979).
Likewise, transfer is an important operation in a Hypothesis Testing model, where it constitutes one possibility for the learner to
establish hypotheses about L2 rules. By contrast, transfer does not contribute to learning at all in the Monitor Model, which
regards L2 learning as being exclusively brought about by an interaction of L2 input with a 'creative construction' mechanism in
the learner. According to this model, transfer only occurs as a production procedure, activated when L2 knowledge is not yet
sufficiently available. Finally, the Interaction Approach, which sees L2 learning as an outcome of the learner's communicative
interaction with other native or non-native speakers, has nothing to say about the contribution of different (linguistic) knowledge
sources to L2 learning.
To take an analogy from a related area: just as the notion of communication strategy has to be, and has been, related to explicit
models of communication (i.e. models of speech production, reception and interaction) learning strategies derive their status as
scientific concepts only from an explicit theory of L2 learning.
Now the mental and interactional behaviours that learners themselves perceive as learning strategies provide interesting insights
into students' implicit assumptions about language learning. As these assumptions are likely to function as action-directing
cognitions for learners, it is an important task for L2 research to explore such cognitions (e.g. Wenden, 1984). Yet I cannot see
any grounds on which it would seem justifiable to move directly from such phenomenological descriptions to educational
recommendations and instructions. We need a theory to interpret the learners' reports, and on which educational applications can
be based. Among other things, the theory would have to predict which learning activities are consciously controllable. Evidently,
much L2 learning must be occurring without the learner being consciously aware of it. It is mandatory to identify the conditions
under which learning takes place with more or less attention, so that learning strategies can be directed to areas which are
accessible to, and profit from, conscious manipulation.

3
Research Methods in Empirical Studies of Learning Strategies
Just as our categorisation of phenomena is theory-dependent, the phenomena accessible to observation depend on the methods of
data collection. It is interesting to note that in the descriptive ESL study, neither classroom observation nor interviews of
teachers proved to be particularly informative, but that the most successful method involved eliciting information directly from
the students themselves. Verbal reports by learners have yielded valuable information about learners' cognitive processes in a
variety of L2 tasks, e.g., reading, writing, translation, test-taking, solving reception and production problems, and learning
strategies (cf. Færch & Kasper, 1987).

< previous page page_40 next page >


< previous page page_41 next page >
Page 41
Yet the validity of these data depends crucially upon the type of introspective method employed. In the descriptive ESL and FL
studies, learners are asked to self-report on their learning operations, i.e. to produce generalised statements about how they
usually tackle a learning task. Such statements reveal more about the learners' beliefs about how they learn than about how they
actually perform a specific learning task. As there is a temporal lag between the original thought process and the verbalisation,
and as the verbalised memory content, moreover, is likely to conflate several events, one has to be cautious of the elaborations
and reductions that contaminate the memory of the original cognition.
These shortcomings of the descriptive studies are remedied in the ESL listening comprehension and FL longitudinal studies.
These studies work with thinking-aloud protocols, probing simultaneously into the ongoing thought processes while the student
is carrying out the task. According to Ericsson's & Simon's model of cognitive processing (e.g. 1984), thinking-aloud protocols
yield more valid data on two accounts: first, because of the temporal relationship between the verbalisation and the cognitive
processing. The verbalisation takes place while the information is stored in short-term memory, i.e., no further reductive or
elaborative processing has occurred. Secondly, because the verbalization relates to a specific event rather than stating
generalised behaviour. Surely, the interpretation of learning strategies identified on the basis of thinking-aloud data is still
dependent on our model of L2 learning, but thinking-aloud data provide a more valid account of phenomena to start with.

4
Universality/Specificity of Learning Strategies
One important conclusion of Dr Chamot's paper is that ESL and FL students seem to use the same learning strategies. This
suggests that learning strategies have the status of cognitive universals. However, in the learning strategies training study, there
was an interesting cultural difference to be noted. While the Hispanic experimental students (the group having received
instruction in learning strategies) did better than the Hispanic controls in a vocabulary task, Asian controls did better than the
Asian experimental students in the same task. The explanation was that the Asian students were trained to use rote memorisation
very effectively, while having to use different learning strategies proved less adequate for them. Other instances of culturally
determined strategy use come to mind. For example, showing lack of knowledge in or out of the classroom may be experienced
as facethreatening by members of some cultures. Consequently, overt hypothesis testing may be avoided. Likewise, failure to
have understood a teacher's utterance may be experienced as embarrassing for the student and disrespectful to the teacher,
resulting in avoidance of comprehension checks and requests for clarification. It seems to me that as long as we do not have
good evidence for the universality of a learning strategy, we have to be

< previous page page_41 next page >


< previous page page_42 next page >
Page 42
attentive to possible cultural differences, in order to be able to support students in using strategies that are in accordance with
their cultural and individual preferences.

References
Canale, M. and Swain M. (1980) Theoretical bases of communicative approaches to second language teaching and testing.
Applied Linguistics, 1, 1-47.
Cook, V. (1985) Universal grammar and second language learning. Applied Linguistics, 6, 2-18.
Cummins, J. (1979) Cognitive academic language proficiency, linguistic interdependence, the optimum age question and some
other matters. Working Papers on Bilingualism, 19, 121-29.
Ericsson, K.A. and Simon, H.A. (1984) Protocol Analysis. Cambridge, Mass.: MIT.
Færch, C., Haastrup, K. and Phillipson R. (1984) Learner Language and Language Learning. Copenhagen: Gyldendal and
Clevedon: Multilingual Matters.
Færch, C. and Kasper, G. (1985) Procedural knowledge as a component of foreign language learners' communicative
competence. In H. Bolte & W. Herrlitz (eds), Kommunikation im Sprachunterricht. Utrecht: Rijksuniversiteit, 169-99.
Færch, C. and Kasper, G. (eds.) (1987) Introspection in Second Language Research. Clevedon, England: Multilingual Matters.
Gass, S. (1979) Language transfer and universal grammatical relations. Language Learning, 29, 327-44.
Hatch, E. (1984) Theoretical review of discourse and interlanguage. In A. Davies, C. Criper and A.P.R. Howatt (eds),
Interlanguage. Edinburgh: University Press, 190-203.
Krashen, S.D. (1981) Second Language Acquisition and Second Language Learning. Oxford: Pergamon.
Martin-Jones, M. and Romaine, S. (1986) Semilingualism: a half-baked theory of communicative competence. Applied
Linguistics, 7, 26-38.
Naiman, N., Fröhlich, M., Stern, H.H. and Todesco A. (1978) The Good Language Learner (= Research in Education Series No.
7). Toronto: OISE.
O'Malley, M., Chamot, A.U., Stewner-Manzanares, G., Kupper, L. and Russo, R.P. (1985) Learning strategies used by
beginning and intermediate ESL students. Language Learning, 35, 21-46.
Rubin, J. (1981). Study of cognitive processes in second language learning. Applied Linguistics, 2, 117-31.
Wenden, A.L. (1984) Theories of second language learners: Making sense of their strategies. Paper presented at the 18th Annual
TESOL Convention, Houston, Texas.

< previous page page_42 next page >


< previous page page_43 next page >
Page 43

The Cognitive Academic Language Learning Approach (CALLA)


J. Michael O'Malley
Evaluation Assistance Center (EAC) East, Georgetown University,
1916 Wilson Boulevard, Suite 302,
Arlington, Virginia 22201, U.S.A.
Abstract. The 'Cognitive Academic Language Learning Approach' (CALLA) is designed for limited English proficient
(LEP) students who are being prepared to participate in mainstream content area instruction. CALLA provides transitional
instruction for upper elementary and secondary students at intermediate and advanced ESL levels. This approach furthers
academic language development in English through content area instruction in science, mathematics and social studies. In
CALLA, students are taught to use learning strategies derived from a cognitive model of learning to assist their
comprehension and retention of both language skills and concepts in the content areas. This paper first discusses the
rationale for CALLA and the theoretical background on which the approach is based. This is followed by a description of
the three components of CALLA: a curriculum correlated to mainstream content subjects, academic language development
activities and learning strategy instruction. Suggestions for integrating these three components in a lesson plan are
presented, followed by a discussion of efforts to implement the approach.
There is dramatic national evidence that limited English proficient (LEP) students in the United States encounter difficulties in
learning English and consequently fall behind their native English speaking peers. The evidence for these educational
difficulties is seen in lower educational achievement, fewer years of schooling completed, higher dropout rates and lower status
occupational attainments (Duran, 1983; Hispanic Policy Development Project (HPDP), 1984; National Center for Education
Statistics (NCES), 1982; Newman, 1978; Office of the Assistant Secretary of Defense (OASD), 1982; Roth, 1982; Veltman,
1980). One conclusion which can be drawn from this evidence is that existing instructional programmes have not been adequate
to meet the demands for LEP students to obtain an adequate measure of educational success.

< previous page page_43 next page >


< previous page page_44 next page >
Page 44
This paper identifies an instructional approach designed for LEP students in U.S. schools which is based on an analysis of the
language needs of the students, the academic demands of educational curricula, and the need for students to develop effective
processes for learning academic information. The approach is intended as a bridge or transition programme between an English
as a second language or a bilingual programme and mainstream instruction. Mainstream instruction is the standard academic
curriculum offered to native English speaking students. The need for a bridge to the mainstream for LEP students is evident,
since many LEP students experience success in their special language classes but have difficulty making the transition from an
ESL classroom to academic programmes provided in mainstream education.
This paper describes the rationale for the instructional approach, gives a brief summary of the theory underlying it, describes the
components of the approach, and discusses its applications in classrooms with LEP students. The paper also identifies the
procedures involved in development of instructional lessons within the approach, and suggests how they function in classrooms.
Finally, the paper will share some information on how the approach can be used and efforts to implement it that are planned.
The major objective of English as a second language (ESL) and bilingual programmes at the elementary and secondary levels in
the United States is to prepare students to function successfully in classrooms where English is the medium of instruction for all
subject areas. In almost all bilingual and other forms of instruction for LEP students, formal ESL instruction is an essential
component. Consequently, virtually all LEP students who receive special services tailored to their language needs receive ESL at
a minimum. Students in ESL programmes are expected to develop important skills in English and may become quite proficient
in day-to-day survival in English. At the conclusion of between one and three years of ESL instruction, minority language
students may perform satisfactorily on language proficiency assessment measures and be judged by their teachers as proficient in
English communicative skills. They are then mainstreamed into the all-English curriculum, where typically they encounter
severe difficulties with the academic programme.
The difficulties LEP students encounter in mainstream classrooms may be attributed in part to three factors. The first of these is
the increased language demands made by the academic curriculum, particularly as students move beyond the primary grade
level. Rather than learning to use English as a second language, these students are required to use the target language to learn
difficult concepts and principles. Instead of simply learning new vocabulary or linguistic structures, they must think in the
language, reason through to conclusions, read expository text, develop arguments, express themselves coherently either orally or
in writing, and analyse, synthesise, or evaluate ideas and concepts. Various researchers have found that the development of these
academic language skills lags behind the development

< previous page page_44 next page >


< previous page page_45 next page >
Page 45
of social communicative language skills, often by as much as five to seven years (Collier, 1987; Cummins, 1983, 1984; Saville-
Troike, 1984).
A second factor that may account for the difficulty LEP students experience is inherent in the nature of the subject matter
courses they must take. Many of the unique English language features used in science, mathematics and social studies courses
are substantially different from the language features used in their ESL courses. The technical vocabulary is unique to each of
the content areas, some linguistic structures are different, and the discourse styles of the three areas are demanding in a different
way from those they have experienced before. Even if they have been exposed to similar concepts in their native languageas in a
bilingual programme, or perhaps in their native countrythey will have difficulty transferring the underlying meaning because of
the unique characteristics of English and the way it is used in the content areas. This difficulty becomes particularly acute at the
middle and upper grade levels, where the language used in content areas becomes increasingly academic and less closely related
to the language of everyday communication than is the case at the primary level.
A third possible factor that may lead to achievement problems among LEP students is the need to use effective strategies for
learning academic content while at the same time meeting the demand to learn and use academic language in English. The
attention and mental processing capacity of LEP students is consumed with attempting to comprehend and produce the complex
language required in the mainstream. These students have little attention or mental processing capability remaining to figure out
effective strategies that will help them in learning or retrieving academic content. Native English speaking students have an
accumulated exposure to grade-appropriate academic content from the first grade forward and have many of the initial
vocabulary and language structures required to comprehend new information. Consequently, their attention is not focussed on
processes for figuring out the language but more on the academic content itself. With far greater facility than an LEP student,
native English speakers can apply strategies to plan for and monitor the learning process, relate new information to existing
concepts, organise new information in ways that are more effective for their own learning, infer meanings for unfamiliar words
and concepts, and evaluate the success of the learning effort.

The Cognitive Academic Language Learning Approach


The 'Cognitive Academic Language Learning Approach' (CALLA) is an instructional method for LEP students who are being
prepared to participate in mainstream content instruction (Chamot & O'Malley, 1986, 1987). CALLA (pronounced /kalá/)
combines English language development with content-based ESL and with instruction in special learner strategies that will help
students understand and remember important concepts. CALLA is a

< previous page page_45 next page >


< previous page page_46 next page >
Page 46
transition programme that forms a bridge between an ESL or a bilingual programme and mainstream instruction. CALLA is
designed to meet the educational needs of three types of LEP students:
students who have developed social communicative skills through ESL or exposure to an English speaking environment, but
have not developed academic language skills appropriate to their grade level;
students leaving bilingual programmes who need assistance in transferring concepts and skills learned in their native language to
English; and
bilingual English dominant students who are even less academically proficient in their native language than in English and need
to develop academic English language skills.

The bridge that CALLA provides between special language programmes and mainstream education is illustrated in Figure 1.
LEP students in ESL and bilingual programmes develop initial skills in understanding, speaking, reading and writing in English,
and practise the essentials of communication to negotiate social interactions. These students are typically classified as having
intermediate level English language skills. In CALLA, students at the intermediate and advanced levels obtain additional
experience in English language development specifically related to three academic areas: science, mathematics and social
studies. The intent is to introduce vocabulary, structures and functions in English using concepts drawn from content areas.
CALLA is not designed to substitute for mainstream content area instruction, or to teach the total content expertise required in
school district curricula, as may be the intention in other English language programmes.
The three content areas addressed by CALLA can be phased into the intermediate level ESL class one at a time. We recommend
beginning with science as the first content area because, by using a discovery approach to science, teachers can capitalise on
experiential learning opportunities which provide both contextual support and language development. The next subject that can
be introduced in a CALLA programme is mathematics, which has less contextual support and a more restricted language register
than science. Social studies is the third subject introduced in the CALLA model because, of the three, it is the most language
and culture dependent and includes many topics which are not easily adapted to most experiential learning activities. We plan to
add to the CALLA model a fourth subject area, literature and composition, which will be introduced last in the sequence.
The CALLA model has three components: a curriculum correlated with mainstream content areas, English language
development integrated with content subjects, and instruction in the use of learning strategies. Each of these components is
discussed separately after the discussion of the theoretical framework underlying CALLA and how it is applied. The final
section of this paper provides guidelines for integrating the components into a single instructional approach.

< previous page page_46 next page >


< previous page page_47 next page >
Page 47

Figure 1
The CALLA model: a bridge to the mainstream.

Theoretical Background and Applications


The CALLA approach is based on a theory which addresses a number of questions related to the role of cognition in second
language acquisition. The first question concerns the way in which information is stored in memory. How is information stored
in memory, and are the facts and rules found in academic content stored differently from complex cognitive skills such as
language? The second question concerns the mental processes involved in learning new information. What are the processes
involved in learning, and are they any different for different types of information? A third question concerns the influence of
intentional learner strategies on the acquisition of new information. Are there indentifiable types of learning

< previous page page_47 next page >


< previous page page_48 next page >
Page 48
strategies, how do they influence learning, and can the strategies be taught like any other complex skill?
The answer to all three questions and the theoretical rationale for the CALLA approach emerge from an inspection of recent
developments in cognitive psychology. We have taken one of the more prominent theories in cognitive psychology and extended
it to second language acquisition and particularly to CALLA procedures (O'Malley, Chamot & Walker, 1987). The theory is
based largely on the work of John Anderson (1981, 1983, 1985) but incorporates many findings and theoretical developments in
cognitive psychology and information processing in general. In the following sections, I shall briefly describe the theory and
then return to illustrate how the theory addresses each of the three questions and explains the role of CALLA.
In this theory, information is viewed as being stored in memory in one of two ways: as declarative knowledge, or what we know
about a given topic; and procedural knowledge, or what we know how to do. Examples of declarative knowledge include the
definitions of words, facts and rules. This type of knowledge is represented in long-term memory as meaning-based concepts
rather than precisely replicated events or specific language sequences. We store this kind of information as meaning-based
concepts primarily because capacity limitations of short-term memory restrict the number of independent words or language
structures that can be retained at one time. To compensate for these limitations, we parse incoming language or convert it to
meaning-based representations and then store these representations as propositions in long-term memory. The storage is
accomplished by finding interconnected networks of concepts, called schemata, with which the new information can be
associated, based on common meaning.
Procedural knowledge is used to describe complex cognitive skills and underlies our ability to understand and generate
language. In Anderson's theory, procedural knowledge is represented in memory by production systems, which are rule-based
conditional actions (IF-THEN relationships) that are represented initially just like declarative knowledge but may become
automatic through repeated practice. Production systems have been used to describe procedural knowledge in reading,
mathematical problem solving and chess as well as language comprehension and production. Production systems have been used
by Anderson (1983) and MacWhinney & Anderson (1986) to represent linguistic rules, and by O'Malley, Chamot & Walker
(1987) to represent sociolinguistic, discourse and strategic competence. Whereas declarative knowledge or factual information
may be acquired quickly, procedural knowledge such as language skill is acquired gradually and only with extensive
opportunities for practice.
The first question raised above concerned how information is stored in memory. The theory posits two modes by which
information is stored, declarative and procedural knowledge. This allows for a distinction to be made between factual
knowledge, as in most content area information, and complex cognitive skills, such as problem solving, reading and other

< previous page page_48 next page >


< previous page page_49 next page >
Page 49
language-related skills. Procedural knowledge is dynamic in that it involves recall and modification of information rather than
simply accessing or augmenting existing declarative knowledge. Transformations in using procedural knowledge may come
about through application of an operation to existing informationas in summarising, restructuring or analysing textual
information; developing an outline of the information; or in analysing the requirements of a word problem and applying a
mathematical operation to a set of numbers. What is important for instruction is that declarative knowledge is acquired
differently from procedural knowledge in that declarative knowledge is relatively easy to acquire, while procedural knowledge is
difficult to learn and may require extensive opportunities for practice.
Many of the things we know about declarative knowledge are incorporated into the CALLA approach. We know that
declarative knowledge is learned more effectively by relating existing concepts in long-term memory to the concepts students
will be expected to learn. CALLA lesson plans are designed to begin with a preparation phase in which students are asked to
brainstorm what they already know about a new topic as a means of getting these concepts into short-term memory where they
can be inter-related to the new ideas presented in a lesson. This capitalises on existing meanings the student may understand
related to the topic, which enhances the student's efforts to construct new meanings. We also know that declarative knowledge is
learned more effectively when students elaborate on new information by adding thoughts of their own which may be images,
examples, details, logical inferences or implications of the information. In CALLA, students are encouraged to elaborate in this
manner with new learning. These elaborations also help in recall of knowledge because they provide alternative routes by which
the student can access the information in memory. Another condition that is known to assist in learning declarative knowledge is
for the student to organise the information or divide it into subsets and indicate the relationship among the subsets.
Organisational strategies for science and social studies concepts can be especially useful for students with little prior exposure to
academic information and few schema to which new information can be related.
The second question concerned the way in which complex cognitive skills are learned. Declarative knowledge can become
procedural through practice. Declarative knowledge is illustrated when a student articulates the rules for addition with two place
numbers. But this knowledge may become procedural when the student is able to perform the operations automatically with a
minimum of errors. Similarly, in language, a student may articulate the rules for word combinations but only with laborious
effort be able to construct sentences. When knowledge of the rules becomes procedural, the student will be able to combine
words into sentences rapidly and without difficulty.
Students may reach this more automatic production of language in at least two ways. One is through a three-stage process that
begins with a cognitive

< previous page page_49 next page >


< previous page page_50 next page >
Page 50
stage, continues with an associative stage, and ends with an autonomous stage. In the cognitive stage, individuals commit to
memory a set of facts or rules for production of a complex cognitive skill and initiate practice of the skill. In the associative
stage, practice is repeated and errors begin to drop out as speed in performing the activity increases. The outcome of the
associative stage is a successful procedure for performing the skill. At the autonomous stage, the skill is executed even more
rapidly, automatically, and accurately. There is no sharp distinction between the associative and autonomous stages. At the
autonomous stage, the verbal mediation necessary to perform the skill may disappear, and recall of the rules governing the
action may even be lost. This stage is characterised by tuning the skill to perfection.
Simple repetition of rules during the cognitive stage may not be the most effective way to develop smoothly executed procedural
skills (Gagne, 1985). Most individuals find it aversive to look back repeatedly at a list or a rule to guide them in the next step.
As Gagne (1985: 132) notes, 'there is a common strategy that supports the learning of any procedural knowledge: practice
followed by feedback'. That is, effective acquisition of procedural skills might better involve opportunities to practise the
procedure with prompts to cue the direction of steps to follow. Gagne also points out the importance of generalisation, so the
student will be able to extend the conditions under which the procedure can be performed, and of discrimination, or recognising
the limitations on the range of conditions for performing the action. In addition, individuals may use composition to combine
related productions that have become active in working memory simultaneously.
As with declarative knowledge, many aspects of what we know about learning procedural knowledge are incorporated into the
CALLA approach. In CALLA, students have ample opportunities for practising all complex procedural skills, including problem
solving routines and academic language. Students intially practise procedural skills under structured conditions with cued
prompts for correct responding. Typically, a procedural skill might be segmented so that parts of it can become routinised before
the full skill is exercised through composition. In later stages of learning, the cued prompts are withdrawn so students can
become accustomed to performing the skill in complete form. Discrimination cues are provided so that students know when to
perform the skill and under which conditions the skill is appropriate. Students are also taught to generalise use of complex skills
to other tasks and settings where they can be used.
With language skills, students have opportunities to practise comprehension and production in situations where feedback focuses
on the ability to convey meaning rather than accuracy of grammatical structure or pronunciation. In this way, students do not
become burdened at the onset with rules for production that would delay performance of the skill. With problem-solving skills,
students are carefully guided through the steps of a problem-

< previous page page_50 next page >


< previous page page_51 next page >
Page 51
solving strategy which the student then composes into sequenced steps as the importance of individual steps in problem solution
is understood.
The third question concerned the influence of intentional learner strategies on the acquisition of new information. Extensive
evidence supports the notion that effective learners use different kinds of mental processes for learning from those of ineffective
learners, or at least use the strategies in different ways for specific tasks. Furthermore, we know that strategies can be described
and classified, they can be taught, they will generalise to tasks other than the ones on which they are initially learned, and they
are especially effective in learning academic content and academic language skills (Derry & Murphy, 1986; O'Malley et al.
1985a; O'Malley et al., 1985b; Weinstein & Mayer, 1986).
In CALLA, learning strategies are the basis for the instructional approach. Learning strategies are seen as an example of
procedural knowledge and consequently are difficult to learn and require extensive opportunities for practice. However, like all
procedural skills, they can be learned most effectively using a combination of cued rehearsal with discrimination and
generalisation procedures coupled with opportunities to compose complete strategy performance from smaller components of the
strategy. Students taught through CALLA are provided with guided instruction on the strategy application as is necessary for the
strategy to become proceduralised and generalised. This may include cued prompts for strategy use in a student workbook,
embedding the strategies systematically throughout a workbook so they become an automatic step to follow in completing
assignments, direct instruction on the strategy use and applications, and self-evaluation by the student on the success with which
the strategy aids learning. Learning strategies are applied in CALLA both to English language development and to content area
instruction.

The Content-Based Curriculum


The purpose of CALLA is to provide a broad framework for using language to learn through the integration of language and
content. Content-based English language development is not only important for developing academic language skills, but is also
inherently more interesting to many students than ESL classes, which focus on language only (Mohan, 1986). Content areas
such as science, mathematics and social studies present numerous topics related to a variety of personal interests, and LEP
students can be motivated not only by the topics presented but also by knowing that they are developing the concepts and skills
associated with these subjectsin other words that they are actually doing 'real' school work instead of merely learning a second
language for applications that have yet to be revealed.
LEP students making the transition from a special language programme, such as bilingual education or ESL, need systematic and
extensive instruc-

< previous page page_51 next page >


< previous page page_52 next page >
Page 52
tion and practice in the types of activities they will encounter in the mainstream class. An occasional, randomly selected ESL
lesson on a topic in social studies or science will not adequately prepare students for the type of language-related activities in
these subjects that they will encounter in the mainstream classroom. This is especially true in the middle and upper grade levels,
where the curriculum in the content areas becomes progressively more demanding, both in terms of information load and
language demands. To be most effective, a content-based ESL curriculum should encompass the sequence and major scope
areas of the mainstream curriculum. The topics incorporated should be authentic and important topics for the grade level of the
student.
The CALLA model ensures that the content component is based on the mainstream curriculum for the grade level of the
students who will participate in the programme. This does not mean that it should be identical to the mainstream programme or
that it should replace the mainstream programme. Rather, CALLA includes a sample of high priority content topics that develop
academic language skills appropriate to the subject area at the student's grade level. To select content topics for CALLA lessons,
ESL teachers can co-ordinate with classroom teachers and consult subject area textbooks for the grade level concerned.
Classroom teachers can identify the most important concepts and skills taught in the content areas they teach. Science,
mathematics and social studies textbooks can be used as a source of specific information to be presented. Having used these
resources to identify lesson topics, the ESL teacher can build language development activities onto the content information
selected.

English Language Development


The second component of the CALLA model is English language development. The purpose of this language development
component is to provide students with practice in using English as a tool for learning academic subject matter. Reading and
language arts can be taught as part of content area subjects such as science, mathematics and social studies. The language
demands of the different content subjects, which include the language of curriculum materials and of classroom participation,
need to be analysed so that students can be taught the actual language functions, structures and subject-specific vocabulary that
they will need to enter mainstream content areas.
The language demands of the content areas are different from the typical demands for language skills in ESL classrooms. These
differences can be analysed along the two dimensions suggested by Cummins (1982, 1983) to represent language proficiency.
The first dimension concerns contextual cues that assist comprehension and the second concerns the complexity of the task.
Language that is most comprehensible is contextualised and rich in non-verbal cues such as concrete objects, gestures, facial
expressions and

< previous page page_52 next page >


< previous page page_53 next page >
Page 53
visual aids. Language that is least comprehensible is language in which context clues have been reduced to such a degree that
comprehension depends entirely on the listener's or reader's ability to extract meaning from text without assistance from a non-
verbal context. The second dimension used to describe the language demands LEP students encounter, task complexity, suggests
that comprehension is affected by the cognitive demands of the task. Language tasks can range from demanding to
undemanding. Examples of less demanding language tasks are vocabulary, grammar drills and following directions. The more
cognitively demanding tasks call on integrative skills such as reading and listening comprehension, and speaking or writing
about academic topics, especially when reasoning skills are required. By combining these two dimensions, tasks involving
language use can be classified into one of four categories: easy and contextualised, difficult but contextualised, context-reduced
but easy, and context-reduced as well as difficult.
Figure 2 is based on Cummins' (1982, 1983) model of two intersecting continua for second language tasks. Allen (1985) has
pointed out that ESL

Figure 2
Classification of language and content activities (adapted from Cummins, 1983)

< previous page page_53 next page >


< previous page page_54 next page >
Page 54
classes generally stress activities in Quadrant I, and that students are then mainstreamed directly into Quadrant IV activities. The
sample activities listed in each quadrant can provide the teacher with information about activities appropriate to the English
proficiency level and age or grade placement of LEP students. The activities listed in Quadrant I might be used in beginning
level ESL classes. LEP students' first experiences with decontextualised language could be planned around the activities listed in
Quadrant II. Some of these activities relate to personal or social communication, and others relate to school activities involving
mainly rote learning. Academic content is included in the activities listed in Quadrant III, but context needs to be built into the
activities to assist comprehension. Activities in this quadrant require hands-on experience and concrete referents. Finally, the
activities listed in Quadrant IV represent those of the mainstream classroom at the upper elementary and secondary levels.
These are the kinds of activities that LEP students have most difficulty with because they are cognitively demanding and
because the language associated with them is reduced in context. Although ESL students probably need to begin with Quadrant I
activities and must eventually be able to perform Quadrant IV activities, some teachers may prefer for activities in Quadrants II
and III to take place concurrently rather than sequentially.
CALLA provides students with extensive transitional language activities in Quadrant III and gradually initiates some practice
with the context-reduced and cognitively demanding activities of Quadrant IV. LEP students need to develop not only content
area reading skills but also the listening, speaking and writing skills associated with each subject. Thus, the CALLA approach
ensures that all four language skills are represented throughout the curriculum in each content area.

Learning Strategy Instruction


The CALLA model uses learning strategy instruction as an approach to teaching the content-based language development
curriculum described in the preceding sections. Learning strategy instruction is a learner-oriented approach to teaching that helps
students learn conscious processes and techniques that facilitate the comprehension, acquisition and retention of new skills and
concepts (Dansereau, 1985; Rabinowitz & Chi, 1987; Rigney, 1978). The use of learning strategy instruction in second language
learning is based on four main propositions (see Cohen & Aphek, 1981; Derry & Murphy, 1986; Gagne, 1985; Naiman et al.,
1978; Rubin, 1981; Segal, Chipman & Glaser, 1985; Weinstein & Mayer, 1986; Wenden, 1985):
1. Mentally active learners are better learners. Students who organise new information and consciously relate it to existing
knowledge should have more cognitive linkages to assist comprehension and recall than do students who approach each new
learning task as something to be memorised by rote learning.

< previous page page_54 next page >


< previous page page_55 next page >
Page 55
2. Strategies can be taught. Students who are taught to use strategies and provided with sufficient practice in using them will
learn more effectively than students who have had no experience with learning strategies.
3. Learning strategies transfer to new tasks. Once students have become accustomed to using learning strategies, they will use
them on new tasks that are similar to the learning activities on which they were initially trained.
4. Academic language learning is more effective with learning strategies. Academic language learning among students of
English as a second language is governed by some of the same principles that govern reading and problem solving among native
English speakers.
While research evidence supports the first two propositions, the transfer of strategies to new learning requires extensive
instructional support. The discriminations and generalisations noted by Gagne (1985) that are required for learning procedural
knowledge apply in this instance since learning strategies are represented in memory by production systems. As noted
previously, CALLA makes learning strategies an integral part of every lesson, provides faded cues to support strategy use, and
encourages the teacher to provide direct instruction in the use of strategies or the explicit explanation of the purpose of the
strategy and how it is used on specific tasks. The fourth proposition is based in part on our own observation that strategies for
language learning are similar to strategies for learning content (O'Malley et al., 1985a) and in part on our positive experience in
training learning strategies on integrative language tasks among ESL students (O'Malley et al., 1985b). While strategies for
learning language and content are very similar, LEP students may have difficulty in discovering the application of strategies to
content because their attention is focussed on figuring out and comprehending the English language. In CALLA, the blending of
language and content results in unified learning tasks to which students can apply learning strategies that facilitate the
comprehension and retention of both declarative and procedural knowledge.
In our own work (O'Malley et al., 1985a) we have identified three major categories of learning strategies:
1. Metacognitive strategies, which involve executive processes in planning for learning, monitoring one's comprehension and
production, and evaluating how well one has achieved a learning objective.
2. Cognitive strategies, in which the learner interacts with the material to be learned by manipulating it mentally (as in making
mental images, or relating new information to previously acquired concepts or skills), or physically (as in grouping items to be
learned in meaningful categories, or taking notes on or making summaries of important information to be remembered).
3. Social Affective strategies, in which the learner either interacts with another person in order to assist learning, as in co-
operation or asking

< previous page page_55 next page >


< previous page page_56 next page >
Page 56
questions for clarification, or uses some kind of affective control to assist a learning task.
We have selected a smaller number of strategies from the original list of those reported by ESL students (see O'Malley et al.,
1985a) for teachers to use as the principal instructional approach for CALLA. These, we believe, are strategies which are easy
to teach, useful for both discrete and integrative language tasks, and applicable to both content and language learning. Some
learning strategies are particularly powerful because they can be used for many different types of learning activities. For
example, metacognitive strategies which can be applied to any type of learning are selective attention, self-monitoring and self-
evaluation. Students can use selective attention to assist comprehension by attending to the linguistic markers that signal type of
information that will follow. Self-monitoring is a metacognitive strategy which has been linked to productive language, in which
students correct themselves during speaking or writing. We have discovered that effective ESL listeners also use self-monitoring
to check on how well they are comprehending on oral text (O'Malley, Chamot & Kupper, 1986). The consequence of self-
monitoring is that students have more time in which they are actively involved in the comprehension and learning task. Self-
evaluation assists learning by helping students decide how well they have accomplished a learning task and whether they need to
re-learn or review any aspects of it.
A number of cognitive learning strategies can be used by ESL students to assist learning. Elaboration is one of the most
powerful strategies and can be applied to all four language skills and to all types of content. When students elaborate, they recall
prior knowledge, consciously interrelate parts of what they are learning, and integrate new information to their existing
knowledge structure. School tasks often require students to learn new information on topics where they do not possess existing
knowledge and, consequently, where elaboration may not be possible. It is important in this case to organise the new
information effectively so that the information can be retrieved and future efforts to use elaboration will be fruitful. Grouping is
a strategy which students without prior knowledge on a topic can use to organise or classify the new information so that a
network or schema is established that will make the new knowledge accessible in the future. Grouping is particularly important
in science and social studies content areas, where students need to understand classification systems and cause and effect
relationships. To facilitate comprehension of groups or sets, students may visualise inter-relationships of concepts using the
strategy of imagery as a way of making mental or actual diagrams of the structure of new information. Grouping and imagery
may be particularly effective in conveying knowledge structures to students with interrupted prior education.
Social/affective learning strategies can be helpful for many types of learning activities. Co-operation is a strategy which has
been shown to have positive effects on both attitude and learning. It is particularly useful for

< previous page page_56 next page >


< previous page page_57 next page >
Page 57
LEP students because by working co-operatively on a task, students practise using language skills directly related to an
academic task. Questioning for clarification is also important because students need to learn how to ask questions when they do
not understand. Some LEP students may not know how or when to ask appropriate questions, or even that U.S. teachers expect
students to ask questions. Self-talk is an affective strategy in which students talk to themselves in order to allay anxiety by
reassuring themselves about their own abilities. It has been used as a way of helping students overcome test anxiety, and could
be used in any situation in which students feel anxious about a learning task.
Teachers can help their LEP students become more effective learners in general by showing them how to apply a variety of
learning strategies to different activities that they may encounter in learning English as well as other subjects in the curriculum.
Suggestions for learning strategy instruction include showing students how to apply the strategies, suggesting a variety of
different strategies for the language and content tasks of the curriculum, and providing many examples of learning strategies
throughout the curriculum so that students will be able to generalise them to new learning activities in other classes and even
outside the classroom. Effective transfer of strategies to other classes also requires that students be made aware of the strategies
they are using and be able to verbalise the conditions under which strategies can be used.

CALLA Lesson Plan Model


In order to integrate the three components of CALLA into an instructional plan, we have developed a lesson plan model that
incorporates content area topics, language development activities, and learning strategy instruction (Chamot & O'Malley, 1986,
1987). In this plan, learning strategy instruction is embedded into daily lessons so that it becomes an integral part of the regular
class routine, rather than a supplementary activity. In this way, students have opportunities to practise the strategies on actual
lessons, and use of the strategies becomes part of the class requirements. At first, the teacher shows the students how to use the
strategies and provides reminders and cues so that they will be used. Later, teachers should diminish the reminders to allow
students to use strategies independently.
CALLA lessons identify both language and content objectives, so that teachers can specify both procedural and declarative tasks
for their ESL students. The lessons are divided into five phases:
preparation, in which teachers provide advance organisers about the lesson and students identify what they already know about
a topic, using elaboration as a strategy;
presentation, where teachers provide new information to students, using techniques which make their input comprehensible, and
showing

< previous page page_57 next page >


< previous page page_58 next page >
Page 58
students how to use strategies such as advance organisers and encouraging use of selective attention, self-monitoring,
inferencing, summarising and transfer;
practice, where students engage in applications of information from the presentation phase, often in co-operative learning
settings, and use strategies such as grouping, imagery, inferencing and asking questions for clarification;
evaluation, in which students reflect on their own learning successes and review difficult and easy portions of the task; and
expansion, where students are provided opportunities to relate and apply the new information to their own lives, or to use
thinking skills in anticipating extensions of the information they have learned.

This lesson plan model can be used in adapting existing lessons for LEP students or in developing new lesson plans. Teachers in
CALLA workshops we have presented have applied the model to lessons from their own texts or in other cases have generated
lessons on topics they select in small groups organised by grade level. The lesson plan model has been applied in science,
mathematics and social studies at elementary, intermediate and secondary grade levels. As noted earlier, in the future we shall
include literature and composition as one of the areas in the CALLA approach.

Conclusions
The Cognitive Academic Language Learning Approach (CALLA) was developed as an extension of either ESL or bilingual
classrooms in which students gain exposure to the academic language required in mainstream classrooms. The approach
integrates language development with academic content and uses a learning strategies instructional procedure in which students
are taught to use special strategies that will assist their learning. CALLA is based on a cognitive theory which differentiates
between declarative knowledge, as in the facts and rules of academic content, and procedural knowledge, or the routines and
processes which become automatic with practice. Special approaches that are appropriate for learning declarative and procedural
knowledge are incorporated into CALLA as regular parts of instruction. CALLA is designed for students at the intermediate to
advanced levels of proficiency in the target language and focusses on the academic language skills these students need for
success in content areas.
We have conducted numerous workshops on the instructional rationale and procedures and have met with enthusiastic interest
among teachers, administrators and teacher trainers. We have also described the approach in a monograph, available from the
National Clearinghouse for Bilingual Education, in numerous presentations and articles, and shall describe the approach in a
forthcoming book. A number of school districts are implementing the approach or portions of it and others have expressed
interest in

< previous page page_58 next page >


< previous page page_59 next page >
Page 59
doing so. We are following these attempts to implement the approach with considerable interest and are encouraging involved
school districts to conduct and report their evaluations.

References
Allen, V. (1985) Language experience techniques for teaching subject content to elementary level limited English proficient
students. Paper presented at the Second Annual LAU Center Conference, Columbus, Ohio.
Anderson, J.R. (ed.) (1981) Cognitive skills and their acquisition. Hillsdalc, NJ: Erlbaum.
(1983) The Architecture of Cognition. Cambridge, MA: Harvard University Press.
(1985) Cognitive Psychology and Its Implications. (second edn) New York: W.H. Freeman.
Chamot, A.U. and O'Malley, J.M. (1986) Language learning strategies for children. The Language Teacher, 10, No. 1, 9-12.
(1987) The Cognitive Academic Language Learning Approach: A bridge to the mainstream. TESOL Quarterly, 19, No. 3, 557-
84.
Cohen, A.D. and Aphek, E. (1981) Easifying second language learning. Studies in Second Language Acquisition, 3, No. 2, 221-
36
Collier, V.P. (1987) Age and rate of acquisition of cognitive-academic second language proficiency. Paper presented at the
annual meetings of the American Educational Research Association, Washington, DC.
Cummins, J. (1982) The role of primary language development in promoting educational success for language minority
students. Schooling and Language Minority Students: A Theoretical Framework. Los Angeles, CA: Evaluation, Dissemination,
and Assessment Center, California State University, Los Angeles.
(1983) Conceptual and linguistic foundations of language assessment. In S.S. Seidner (ed.), Issues of Language Assessment. Vol
II: Language Assessment and Curriculum Planning. Wheaton, MD: National Clearinghouse for Bilingual Education.
(1984) Bilingualism and Special Education: Issues in Assessment and Pedagogy. Clevedon, England: Multilingual Matters.
Dansereau, D.F. (1985) Learning strategy research. In J.W. Segal, S.F. Chipman and R. Glaser (eds), Thinking and Learning
Skills: Relating Learning to Basic Research. Vol. I. Hillsdale, NJ: Erlbaum.
Derry, S.J. and Murphy, D.A. (1986) Designing systems that train learning ability: From theory to practice. Review of
Educational Research, 56, No. 1, 1-39.
Duran, R.P. (1983) Hispanics' Education and Background. New York: College Entrance Examination Board.
Gagne, E. (1985) The Cognitive Psychology of School Learning. Boston, MA: Little, Brown.
Hispanic Policy Development Project (HPDP). (1984) Make Something Happen. Vols 1 and 2. Washington, DC.
MacWhinney, B. and Anderson, J. (1986) The acquisition of grammar. In I. Gopnik and M. Gopnik (eds) From Models to
Modules. Norwood, NJ: Ablex.
Mohan, B.A. (1986) Language and Content. Reading, MA: Addison-Wesley.
Naiman, N., Fröhlich, M., Stern, H.H. and Todesco, A. (1978) The Good Language Learner. Ontario, Canada: Ontario Institute
for Studies in Education.
National Center for Education Statistics (NCES) (1982) Hispanic students in American high schools: Background characteristics
and achievement. Bulletin 82-228. Washington, DC: Department of Education.
Newman, M.J. (1978) A profile of Hispanics in the U.S. work force. The Labor Review, December.
Office of the Assistant Secretary of Defense (OASD) (1982) Profile of American Youth. Washington, DC: Department of
Defense.

< previous page page_59 next page >


< previous page page_60 next page >
Page 60
O'Malley, J.M., Chamot, A.U., Stewner-Manzanares, G., Kupper, L. and Russo, R. (1985a) Learning strategies used by
beginning and intermediate ESL students. Language Learning, 35, No. 1, 21-46.
O'Malley, J.M., Chamot, A.U., Stewner-Manzanares, G., Russo, R. and Kupper, L. (1985b) Learning strategy applications with
students of English as a second language. TESOL Quarterly 19, No.3, 557-84.
O'Malley, J.M., Chamot, A.U. and Kupper, L. (1986) The Role of Learning Strategies and Cognition in Second Language
Acquisition: A study of strategies for listening comprehension used by students of English as a second language. Rosslyn, VA:
InterAmerica Research Associates.
O'Malley, J.M., Chamot, A.U. and Walker, C. (1987) The role of cognition in second language acquisition. Studies in Second
Language Acquisition, 9, No. 3.
Rabinowitz, M. and Chi, M.T. (1987) An interactive model of strategy processing. In S.J. Ceci (ed.), Handbook of Cognitive,
Social, and Neuropsychological Aspects of Learning Disabilities. Hillsdale, NJ: Erlbaum.
Rigney, J. (1978) Learning strategies: A theoretical perspective. In H.F. O'Neil (ed.) Learning Strategies. New York: Academic
Press.
Roth, D.M. (1982) Hispanics in the U.S. labor force: A brief examination. Report No. 82-122E. Washington, DC: Congressional
Research Service, Library of Congress.
Rubin, J. (1981) Study of cognitive processes in second language learning. Applied Linguistics, 11, No. 2, 117-31.
Saville-Troike, M. (1984) What really matters in second language learning for academic achievement? TESOL Quarterly, 18,
No. 2, 199-219.
Segal, J.W., Chipman, S.F. and Glaser, R. (eds) (1985) Thinking and Learning Skills: Relating Learning to Basic Research. Vols
I and 2. Hillsdale, NJ: Erlbaum.
Weinstein, C.E. and Mayer, R.E. (1986) The teaching of learning strategies. In M.C. Wittrock (ed.), Handbook of Research on
teaching. (third edition). New York: Macmillan.
Wenden, A. (1985) Learner strategies. TESOL Newsletter, 9, No. 5.
Veltman, C. (1983) Language Shift in the United States. New York: Mouton.

< previous page page_60 next page >


< previous page page_61 next page >
Page 61

Social and Intellectual Functions of Language:


A Fruitful Distinction?
Hans Vejleskov
The Royal Danish School of Educational Studies, 101 Emdrupvej,
DK-2400 Copenhagen NV, Denmark.
Abstract. When the functions of language are discussed, it must be made clear whether one considers (A) the language
system, (B) conversations or (C) utterances. A study of utterance functions in pre-school children is briefly presented in
order to show that within a certain conversation, whether it be 'social' or 'intellectual', the children may well make
utterances with 'social' as well as 'intellectual' functions. Attention is drawn to the activity of role play which often implies
conversations that take place at different levels, and must often be defined 'social-cognitive'. Although it is natural, from a
psychologically-pragmatic point of view, to distinguish between intellectual and social utterances, discrimination between
intellectual and social conversations, at least among pre-school children, is very dubious. In conclusion, the value of the
observation of utterance functions during conversations to the understanding of monolingual as well as bilingual children
is illustrated.

Language, Conversations and Utterances


In the science of psychology, language has traditionally been regarded partly as a means for communication, mediation or social
interaction, and partly as a means for reference, representation or intellectual conceptualisation. The former view stresses the
fact that, by means of utterances, we are able to influence other people, whereas the latter underlines the fact that words name
and substitute things or events. Correspondingly, the term 'meaning' sometimes refers to the intentions on the part of the speaker,
and at other times it refers to the conceptual content of a word or a sentence used by the speaker.
Almost everybody will insist that there need not be any contradiction here, as the two views only deal with the two sides of one
coin. However, the relevance of this 'compromise' depends on one's approach. To be brief, only two different distinctions will be
mentioned:

< previous page page_61 next page >


< previous page page_62 next page >
Page 62
1. a psychologist may look at language as a tool for psychological investigation. For instance, he may consider the very
problem of psychological description and ask how valid are a subject's reports about private experiences;
2. or he may study other persons' language in terms of verbal behaviour, social interaction, etc.

The other distinction is concerned with the meaning of the term 'language'.
A. A psychologist may ask whether the language system is primarily a means for social interaction, or a means for intellectual
performance.
B. He may ask whether the acquisition of language is primarily related to the social or the cognitive development of young
children.
C. Or he may ask whether a certain utterance made in a certain situation has a social or an intellectual function.

The study that is briefly presented in the next section focusses upon point 2 and point C.
In a recent paper, Dabbs (1985) introduces a level between A (language) and C (utterance) as he distinguishes between social
and intellectual conversations. He assumes that each of the two kinds of conversation has emerged from earlier systems,
'. . . one evolving primitive vocal stimuli that convey information about recurring social themes, and the other involving a
grammar based upon strategies the intellect has found useful in dealing with the world' (p. 185).
and he presents some studies which show that the two kinds of conversation are different with regard to facial expressions,
pauses, turn length, eye contact, etc.
The value of Dabb's distinction to the verbal interaction of pre-school children is questioned by our observations.

Utterance Functions in Young Children


In our studies on utterance functions in small groups of pre-school children the following terminology was used: each of the
many different episodes was characterised according to (1) participants (number and age of children, adult participation, etc.)
and (2) kind of activity (role play, construction play, having lunch, etc.). In the various episodes, each utterance was
characterised according to function: on the basis of the utterance said as well as the whole situational context we attempted to
understand what the speaker intended by making that utterance. The 'function of the utterance' is thus roughly synonymous with
'speaker intention' or 'speech act'. For each utterance we also recorded who was the speaker and who was the addressee, who if
any did answer it, and a number of other variables (Vedeler and Vejleskov, 1986).

< previous page page_62 next page >


< previous page page_63 next page >
Page 63
It is obvious that the very definitions of the terms of utterance and episode give rise to problems. How much can the play
change its character before we find that one episode has ended and another episode has begun? How do we decide whether a
certain speech event of a child consists of one, two or three utterances? The latter question can be answered in this way: a
speech event that may consist of one or more (grammatical) sentences, is counted as one utterance if it has only one function,
whereas it constitutes two utterances if the first part is one speech act and the last part another speech act. These questions, as
well as some problems connected with the concept of 'intention', are discussed elsewhere (Vejleskov, forthcoming).
The categorisation of utterance functions used in the interpretation and recording of the video-taped episodes is presented in
Table 1, which also indicates six groups of utterance functions: (A) intellectual, (B) social, (C)
Table 1. Twenty-one utterance functions and a division into six
groups, A-F
Speaker wants to Group
'Examine' listener; to control his knowledge
Get knowledge about something from listener A
Tell listener about something
State a fact, e.g. by answering 'yes' or 'no'
Carry on the conversation, to be polite
Stick to his own; to hold his own B
Ask for support or acceptance
Ask for sympathy
Make contact with listener
Influence another person's feelings positively
Express his own feelings C
Express his own feelings negatively
Influence another person's feelings negatively
Instruct listener; to explain something
Direct other's activities for educational reasons D
Have something done; to get help
Get something E
Direct other's activity ('you')
Direct or plan the activity of his own group ('we') F
Direct or plan his own activity ('I')
No special function

< previous page page_63 next page >


< previous page page_64 next page >
Page 64
emotional, (D) instructional, (E) instrumental and (F) directing. In the present context we focus on groups (A) and (B). The
following observations show that some utterances cannot be characterised as either 'purely intellectual' or 'purely social'.
1. Answering 'yes' or 'no' may be an act of stating a fact as well as an act of politely carrying on the conversation.
2. The intention to 'examine' the listener clearly implies educational utterance functions like those of group (D) and so they are
'social' as well as 'intellectual'.

Nevertheless, the main result of the categorisation and analysis of a total of 17,000 utterances from about 100 different episodes
with about 70 different participants is that it is both possible and natural to attribute the 23 functions to utterances. However, we
often had to characterise the individual utterance by means of a main function and a supplementary function. The analyses have
still not been presented in English, and here we can only briefly mention those results that are specially relevant to the present
discussion:
1. It is often quite natural to attribute a social function to one utterance and an intellectual function to another utterance in the
same episode.
2. Those children who are comparatively the most active speaker-listeners make more utterances with social or intellectual
functions, and fewer utterances with a directing function.
3. In role play episodes even less active speaker-listeners make more utterances with social and intellectual functions
compared with episodes of construction play, cf. Table 2.
4. With small groups of children aged 3-6 years it is not natural to distinguish between episodes with social conversation and
episodes with intellectual conversation as proposed by Dabbs (1985).

In some episodes three children (and their pre-school teacher) were sitting at a table eating lunch, and the verbal interaction
carried out here might naturally be called 'a social conversation'. However, a closer examination revealed that the conversation
changed very frequently: sometimes the children had 'a small talk', sometimes they were talking about themselves, and
sometimes they were 'oriented more towards practical issues or abstract ideas' (Dabbs, 1985: 183). It also turned out that during
all these kinds of 'sub-conversation' utterances with intellectual as well as social functions occurred.

Role Play and Communicative Awareness


In a larger number of the episodes intensively studied by Vedeler (1985), three or four children were role playing, i.e. during a
considerable part of these episodes the children pretended to be another person (a doctor, a

< previous page page_64 next page >


< previous page page_65 next page >
Page 65
Table 2. Percentage distribution of utterances according to group of
function, (cf. Table 1) and according to the existence of an addressee
Vedeler (1985, Tables 13 and 15d)
Type of function Boys aged 3-5 years Boys aged 6 together with
without adult pre-school teacher
Role play Construction play Role play Construction play
Intellectual 12 9 13 19
Social 24 15 27 20
Others 62 69 56 59
No 2 7 4 2
Communicative 66 59 77 79
Private 34 41 23 21
N 499 202 392 287

mother, a hunter, etc.). Accordingly, many utterances were made on the part of such fictional persons, or they dealt with the
behaviour of such persons or the whole imaginative context.
Thus, in this rather common type of play, we observed frequent shifts between three levels: (1) true role utterances, i.e. speech
as role persons; (2) utterances about the role persons; and (3) utterances without any reference to the fictional context.
Obviously the children have to master simultaneously their own interplay and that of the fictional persons. Although they can
stick to their stereotypic conceptions of, for example, medical personnel, they often have to reflect on the behaviour of, for
example, a doctor or a nurse, and they also have to reflect on the content as well as the functions of utterances made by such
persons. In fact, a considerable number of the utterances about the role play are concerned with the use of language ('A doctor
doesn't say so'; 'A nurse has got to be kind').
These observations suggest that the activity of role play gives rise to social cognition as well as communicative awareness in
children.
The issue of social cognition, which is intensively studied in current developmental psychology (cf. Higgins et al., 1983),
includes the perception and understanding of human and social phenomena, i.e. feelings, attitudes, intentions, conventions,
moral values, etc., of oneself as well as other people. So far it has not been connected with the study of the development of
language and uses of language.
By communicative awareness we refer to the growing awareness of verbal interchange and utterance functions, which should be
studied in addition to the study of lingustic awareness.

< previous page page_65 next page >


< previous page page_66 next page >
Page 66
It should be mentioned that the term 'function' is commonly used with two different meanings: 'Language functions' may refer to
the ability to take part in a conversation as one masters various 'technical' functions, e.g. turn-taking. It may also refer to the
speaker's intentions as he produces individual utterances, i.e., performs individual speech acts. Only the latter sense of 'functions'
is considered here. However, the concept of 'communicative awareness' includes both speech acts, conversational techniques,
and social conventions of politeness etc.
Because of the fact that children's role play is directly connected with the phenomena of social cognition and communicative
awareness, it is natural to think that it should be given much attention when children's language acquisition is considered from a
psychological as well as an educational point of view.

The Pragmatic Approach and Bilingualism


The above-mentioned qualities of role play were made clear through efforts to characterise the individual utterances of the
participants in terms of functions or speech acts. In conclusion we put forward the more general assumption that such focussing
on the pragmatic aspects at utterance level is a fruitful approach to children's acquisition of the first as well as the second
language.
Above all, this approach makes it obvious that we should distinguish between the child's functional use of language in a variety
of situations and, on the other hand, the child's formal skills of language, i.e. pronunciation, morphology, syntax, vocabulary,
etc. This distinction may turn out to be more fruitful than that between 'social' and 'intellectual' functions, which, at the level of
conversation and language-system, has often been used in connection with bilingual children, because we, regardless of a child's
formal skills in the second language, will observe and evaluate his efficiency as a second language speaker-listener in different
kinds of situation important for children at that stage of development.

References
Dabbs, J. M. (1985) Temporal patterns of speech. In H. Giles and R. N. St Clair (eds), Recent Advances in Language
Communication and Social Psychology. London: Erlbaum.
Higgins, E. T. et al., (eds) (1983) Social Cognition and Social Development. Cambridge: Cambridge University Press.
Vedeler, L. (1985) Barns kommunikasjon i rolleleik. Oslo: Statens Spesiallærerhøgskole.
Vedeler, L. and Vejleskov, H. (1986) Language in Pre-school children. In M. Leiwo (ed.), Papers from the Fifth Nordic
Conference on Applied Linguistics. Jyväskylä Univ.
Veileskov, H. (forthcoming) Observation of people's intentions. In I. Bjørgen (ed.), Fundamental Problems in Psychological
Research.Oslo: Universitetsforlaget.

< previous page page_66 next page >


< previous page page_67 next page >
Page 67

Lexical Characteristics of Near-Native Second-Language Learners of Swedish 1


Kenneth Hyltenstam
University of Stockholm, Institute of Linguistics, Department of Research on Bilingualism,
S-106 91 Stockholm, Sweden.
Abstract. Various aspects of the vocabulary of near-native secondlanguage learners of Swedish were analysed and
compared to that of matched native Swedish speakers. Quantitative measures of lexical usage focussed upon comprised
lexical density, lexical variation and lexical sophistication. Further, a qualitative and quantitative analysis of lexical errors
was carried out.
Data were elicited from 36 students attending the second-year course of the Swedish gymnasieskola (senior high school
level). The population comprised 24 bilingual students, among whom 12 had Finnish and 12 had Spanish as their first
language, and 12 monolingual Swedish speakers. Both oral and written data were collected from each student, and, from
the bilingual students, data were obtained from both their languages. The oral data consisted of retellings of four
narratives, while the written data comprised summary reviews of Charlie Chaplin's silent film, Modern Times.
Only the Swedish data are considered in this paper. The results are discussed in relation to the current discussion of
differing proficiency levels required for everyday context-embedded, cognitively un-demanding communication as
opposed to literate language use (cf. Cummins, 1983). In addition, the results from the lexical error analysis are related to
the concept of fossilisation (Selinker, 1972).

Introduction
In current literature on bilingualism and second language acquisition there are a number of important theoretical questions that
would need more empirical ground, so that they could be further developed. We shall mention just two examples, taken from
distinct areas, that are of specific interest for the present investigation.
First, we have the theoretical discussion on differing levels of language proficiency, as typically required by oral interactional
language use, on the

< previous page page_67 next page >


< previous page page_68 next page >
Page 68
one hand, and literacy-related language use, on the other. The most influential model of language proficiency, comprising such a
dichotomy, is the one suggested by Cummins (1979). In his terms, the learner can obtain either just basic interpersonal
communicative skills (BICS) or this plus cognitive/academic language proficiency (CALP).
The dichotomy between BICS and CALP has been criticised from many perspectives, among others for its handling of reality in
too simplistic a manner and for its lack of operationalisation (cf. Edelsky et al., 1983; Martin-Jones and Romaine, 1986). Partly
in response to this criticism, but partly irrespective of the debate about Cummins's model, the discussion of proficiency levels
has turned away from dichotomies to dealing explicitly with continua (Brown, 1981; Cummins, 1983). The rather gross
approach to differing language proficiencies in Cummins's particular 1979 model has later been replaced by a two-dimensional
model where two inter-related continua define the demands on language proficiency, namely contextembedded and context-
reduced communication, on the one hand, and cognitively undemanding and cognitively demanding communication, on the
other (Cummins, 1983).
The question of assessing the various levels of linguistic proficiency, however, is as yet unresolved to the extent that the
criticism for lack of operationalisation is valid. Therefore, although it is intuitively appealing and in accord with informal
observations that a certain individual may have difficulties caused by language problems in more demanding communicative
activities, but not in everyday face-to-face interaction, there is still a long way to go before we have found out whether this
observation holds for a more stringent examination, and, if so, what the characteristic linguistic features of the various segments
of the language proficiency continuumor continuamay be.
The second example of notions that are scientifically undeveloped is that of fossilisation which was suggested by Selinker in
1972. Fossilisationaccording to observationsis a process that may occur in the second-language acquisition context as opposed to
first language acquisition. It covers features of the second language learner's interlanguage that deviate from the native speaker
norm and are not developing any further, or deviant features whichalthough seemingly left behindre-emerge in the learner's
speech under certain conditions. Thus, the learner has stopped learning or has reverted to earlier stages of acquisition. Accent in
pronounciation is a typical example of fossilisation. The notion of fossilisation is generally accepted in interlanguage studies,
although there are remarkably few discussions on the nature of the notion in the literature. This is particularly surprising when
one considers how little is understood of the phenomenon as such. A number of questions, such as the following, suggest
themselves:
1. Is fossilisation a feature of second language competence, or is it a processing feature?

< previous page page_68 next page >


< previous page page_69 next page >
Page 69
2. Can fossilised features start to develop anew in later phases of acquisition? (The term itself, of course, would suggest this
should not be the case.)
3. Why does fossilisation occur? Are the reasons for fossilisation dependent on the language faculty as such, or are the causes
to be searched for in the sociopsychological realm only? Here one should note that the 'explanation', sometimes suggested, that
second language learners stop developing their interlanguage as soon as they can make themselves understood may seem very
straightforward, but becomes futile when one realises that first language learners do not stop learning at such a point. Why do
language learners respond to the second-language acquisition task in this way, when children acquiring their first language
bother to go all the way up to the native norm?

Further, among the most intriguing questions are the following two:
4. Does fossilisation occur only in adult learnersor, more precisely, learners who have reached puberty by the onset of the
language learning task? If this were the case, fossilisation must be seen as an age-related language learning phenomenon. If it
can also be found in children acquiring a second language, it could be seen as a pure second language, as opposed to first
language, phenomenon.
5. Are there always some traces of fossilisation left as an outcome of second language acquisition, or, alternatively, as an
outcome of second language acquisition beyond puberty?

Interestingly, 15 years after it was introduced by Selinker, the notion of fossilisation has begun to attract considerable attention.
For example, Kellerman (in press), in discussing certain typical interlanguage features of advanced Dutch learners of English,
tries to predict what type of linguistic features are candidates for fossilisation. Coppieters (1986), in investigating differences in
intuitional judgements of 'peripheral' syntactic structures between native and advanced non-native speakers of French, concludes
that both quantitative and qualitative differences exist in the internalised language systems of the two groups. Schachter
(personal communication), in ongoing research, considers fossilisation features in naturalistic data, and Ioup (1984), in
investigating whether a 'syntactic accent', parallel to the more familiar phonological accent, can be said to exist in advanced
second-language learners, as a matter of fact concludes that it cannot.
Both these areas of interest, the differing proficiency hypothesis and fossilisation, are often discussed against data from the
advanced second-language learner. In the case of proficiency-level differences, learners with native-like command in oral
language proficiency have typically been considered, since what is at stake is precisely that even though a second language
learner may have native-like command in a language, s he may have a level of language proficiency that only allows everyday
face-to-face interaction but not literacy-related language use. In the case of fossilisation, adult learners

< previous page page_69 next page >


< previous page page_70 next page >
Page 70
have been the natural population to consider. In the recent studies mentioned above, where the phenomenon of fossilisation as
such has been in focus, it has been judged profitable to investigate language use in advanced learners, where the differences
from native speakers are presumably limited, and therefore easier to study.

Present Study
Motivation and aim
In this paper we shall consider a subset population among the group of advanced second-language learners, namely what we
prefer to call the near-native second-language learner. What we have in mind is a second language learner who speaks the
second language without any noticeable accent in pronunciation and without apparent deviances in grammar or lexicon. In other
words, near-native learners should pass as native speakers of the second language in everyday conversation.
The reason we identify these speakers as near-native, and not native, is that they may exhibit some linguistic features that
deviate from the native norm, but which typically are difficult to identify as non-native for the listener. It may, for example, be
unclear to the listener if these features are dialectal or non-native. It may happen, sometimes after extended conversation of
hours or days, that it suddenly becomes clear to the listener that non-native features exist. Also, in modalities of language use
other than the oral one, the non-native features may be more obvious.
Thus, investigating linguistic features of the near-native learner may give us insights into whether it is possible to identify
linguistic correlates of the proficiency levels required for different types of communicative activities or other language tasks.
Furthermore, since the near-native learner by definition has mastered almost all aspects of the second language system, the few
deviances from the native norm that do exist should be more certain candidates for inclusion in the category of fossilisation.
Apart from these two motives for studying the near-native learner, there is also an obvious practical pedagogical reason to focus
on this group. When it comes to providing instruction in language skills for this group of learners, the only options offered are
generally either mother tongue instruction in the second language or more basic second-language instruction. There is no body
of knowledge on the specific needs of the very advanced or near-native learner, nor any specific means of dealing with these
needs. Knowledge of the characteristic features of the language of this group is obviously the starting point for creating
supportive pedagogical means for the students.
In this paper we have chosen to present data on lexical aspects of the learners' productions, although this study is part of a more
comprehensive

< previous page page_70 next page >


< previous page page_71 next page >
Page 71
investigation of the language of near-native second-language learners of Swedish (see Hyltenstam and Stroud, in preparation,
for an extensive report). The lexicon is well suited to shed light on some of the questions actualised by the two areas of interest
presented above. The requirements for lexical ability in literacy-related language use are distinct from those posed in everyday
oral face-to-face interaction. Appropriate literacy-related language use is typically characterised by a more varied and a more
specific employment of the lexicon that results in written or spoken texts with higher semantic density than non-literate oral
language use (Ure, 1971; Hultman and Westman, 1977; Biber, 1986; Linnarud, 1986). Thus, one aspect of the proficiency level
necessary for literate language use is the possession of a reasonably large lexicon and the ability to employ it according to the
requirements of literate language use. One of the two questions we would like to address, therefore, in this paper is the
following:
1. Are there any differences between near-native and native speakers in the variation, density and specificity of their lexicon in
literacy-related language use?
As regards the second area of interest, one might expect that the lexicon is particularly sensitive to fossilisation, considering the
fact that the lexicon is more individual in nature than other aspects of language. Even native speakers differ more among
themselves in the content of lexical units than in the value of grammatical or phonological features. Thus, the second question
we address is:
2. Are the near-native speakers different from native speakers in the quality and quantity of lexical units that deviate from the
native norm?
Subjects
The subjects for this study were 36 students, aged 17-18, attending second year courses at the Swedish gymnasieskola (senior
high school level). The subjects comprised 12 speakers that were bilingual in Finnish and Swedish, 12 in Spanish and Swedish,
and 12 monolingual Swedish controls. (The three groups of subjects will be called Finnish, Spanish and Swedish respectively for
short, although this is not quite adequate, given the biculturalism of the bilingual students and the fact that only one of the
'Spanish' subjects comes from Spain; the others are Latin American.) The criteria for selection of the bilingual subjects were (1)
that they should pass as native speakers of Swedish in everyday conversation, (2) that they also used their first language on a
daily basis, i.e. they should be active bilinguals, and (3) they were to represent the whole range of grade levels. The Swedish
subjects were matched with the bilingual subjects for grade level.

< previous page page_71 next page >


< previous page page_72 next page >
Page 72
Task
Each student was invited to participate in a set of oral and written tasks that were all manifestations of literacy-related language.
The written data comprised a summary of a selection of 20 minutes from Charlie Chaplin's silent film Modem Times that had
been shown to the students. The oral data consisted of retellings of four texts (two written and two spoken) that posed different
degrees of cognitive demands on the subjects. One of the written and one of the spoken texts were fables, whereas the other two
were informative expository texts. The subjects read or listened to each text once only, before they had to retell it with the
investigator as the listener.
Analysis
The quantitative measures used to estimate the lexical characteristics of the three groups were lexical density, lexical variation
and lexical sophistication. Lexical density (Ure, 1971; Linnarud, 1986) assesses the proportion of semantically full words as
opposed to function words in a text and gives an estimate of the semantic richness of the text. Lexical variation, which is used in
this study in accordance with Linnarud (1986), gives a measure of the type/token ratio among content words (or lexical words)
in the texts, while lexical sophistication (Linnarud, 1986; cf. also Hultman and Westman, 1977) shows the proportion of
infrequent words in a text. A frequency word list (S. Allén, Nusvensk frekvensordbok, I) based on a corpus of 1,000,000 running
words in newspaper texts, was used as criterion in identifying which words were to be considered sophisticated (words that were
not included among the 7,000 most frequent Swedish words). A high degree of lexical sophistication is characteristic of a text
that is concrete and specific in content (Hultman and Westman, 1977).
The qualitative measure employed in this study was a lexical error analysis. The error categories comprised two main types,
approximations to a target lexical unit, and contaminations of two or more such units. Further, a major division was made
between deviant productions that were single lexical units and those that consisted of lexical phrases. The categories will be
further defined in the results section.

Results
Quantitative measures
The figures for lexical density (LD), lexical variation (LV) and lexical sophistication (LS) in the written and oral data are given
in Table 1. We have chosen to give the group values in tabular form and comment on specific individual results in the text.
Also, the various retellings have been

< previous page page_72 next page >


< previous page page_73 next page >
Page 73
collapsed into one figure for all the oral data. A more detailed account of the results on these measures is given in Hyltenstam
and Stroud (in preparation). The results displayed in Table 1, in fact, are quite unspectacular. The figures for lexical density, i.e.
the proportion of content words of all words, are in complete accord with what has been found in other studies of native
speakers for all our groups. Ure (1971), who studied written and spoken texts in English, found that spoken texts generally had a
lexical density of below 40%, while written texts almost always had values above 40%. Percentages above 40 were also found
by Linnarud (1986) for both native English speakers and Swedish learners of English at the senior high school level (aged 17-
18). Linnarud (1975), on the other hand, found a low LD in 72% of the written texts from the university level students she
studied in that investigation. Her interpretation is that the low values of the university students were due to the content of the
task: writing a composition about an abstract topic.
Table 1. Lexical density (LD), lexical variation (LV) and lexical
sophistication (LS) in written (W) and oral (O) data from Finnish (Fi),
Spanish (Sp) and Swedish (Sw) subjects
LD LV LS
W 0 W 0 W 0
Fi 45 38.3 76.9 61.9 15.7 16.9
Sp 44.3 36.6 67.7 56.2 13.3 13.7
Sw 44.6 35.9 71.9 65.5 15.5 14.6

The differences in LD between the groups in this investigation were all statistically insignificant.
Figures for LV, i.e. the type/token ratio for content words, similarly did not distinguish the groups for the written data. In the
oral data, however, there was a slight advantage for the Swedish group, with a significant difference from the Spanish group.
Interestingly enough, this difference is a result of the Swedish group's higher values on the more cognitively demanding
retelling tasks.
As is well known, type/token ratios are difficult to compare and interpret, since the figures are sensitive to text length. Various
attempts have been made in the literature to overcome this difficulty by finding a formula that eliminates the effect of text
length. In this study, a formula, developed by Tor Hultman (see Hultman and Westman, 1977: 56), has been used, which makes
the measure of variation 'relatively independent of the length'. 2 In other words, the problem is decreased but not eliminated in
our comparisons, which means we cannot draw any firm conclusions on the basis of the difference in LV in the oral data
between the groups.

< previous page page_73 next page >


< previous page page_74 next page >
Page 74
The measures of LS, i.e., the proportion of infrequent content words of all content words, gave similar figures in the written data
for the Finnish and the Swedish groups and slightly lower values for the Spanish group in the written data. On the oral tasks,
the Finnish group had the highest value, followed by the Swedish group, whereas the Spanish group again had a lower value.
Three of these differences were significant: between the Finnish and Spanish group in the written data (0.02 level), and, in the
oral data, between the Spanish and Finnish groups (all levels) and between the Finnish and Swedish groups (0.05 level).
On the whole, we have seen that there are few consistent differences between the groups. The differences that do exist are in
most instances insignificant, and they are difficult to interpret. The results certainly do not distinguish the monolingual from the
bilingual subjects in any consistent way.
Turning now to individual patterns in LD, LV, and LS, we find, of course, pupils with both high and low values on these
measures. The distribution of high values and low values does not, however, coincide with monolingualism and bilingualism
respectively. As a matter of fact, there are four students with consistently high values and none with consistently low values in
the Finnish group. In the Spanish group, there is one subject with consistently high values and two with consistently low values.
Interestingly, in this group we find one subject who at the same time has several quite low values and a number of very high
values. In the Swedish group, finally, there are two students with consistently high values and five with consistently low values.
This means that, in the total population, we have seven subjects with high, seven with low and 22 with average values, a quite
symmetrical distribution. More pupils with high values belong to the Finnish group, whereas more with low values are to be
found in the Swedish group. In the Spanish group, the pupils generally have average values.
One might ask if there is any correlation between values on our measures and grade level in Swedish. In the 'high' group, there
are three students with grade 3, and one each with grades 2, 4 and 5. (The grades given in Swedish schools cover the steps from
1 (lowest) to 5 (highest).) In the 'low' group, there are again three students with grade 3, and two each who have grades 2 and 4.
Thus, unexpectedly, the grade distribution is quite similar in both groups.
This, in fact, is an indication that the measures used may not be sensitive to certain kinds of variation found in our texts. In
working with these measures it soon becomes obvious that they give very rough and sometimes misleading results. It is possible
to obtain high values on all three measures by the wrong means, so to speak: a subject may, for example, get high values on LD
although the vocabulary used is unvaried and small, as pointed out by Linnarud (1975), but also because of the lack of
appropriate grammatical structuring of the text. A telegraphic style, for example, gives very high values for LD. One of the
students in this investigation, who had

< previous page page_74 next page >


< previous page page_75 next page >
Page 75
extremely high values on LD (but low on LV) is a case in point. The fact that he only seldom employs anaphoric
pronominalisation where it would be appropriate accounts to a large extent for his high LD values; instead of using pronouns,
which of course are function words, he repeats the same content words. The following excerpt from this student's written data is
an illustration both of the high LD and the low LV values:
Efter en stund bestämmer After a while the director
chefen att det ska gå fortare decides to increase the speed of
på den löpande banden. Efter the assembly line. After a few
några minuter får en av minutes one of the men who
mannen som jobbar på den works at the assembly line gets
löpande banden rast, så a break, then another man
kommer en annan man och comes and replaces the man.
ersätter mannen. Mannen som The man who gets the break
får rast går på toaletten och goes to the bathroom and
röker. Då helt plötsligt på ena smokes. Then suddenly on one
vägen av toaletten kommer en wall in the bathroom comes the
ansikte av chefen. Chefen face of the director. The
säger till mannen att gå jobba director says to the man to go
igen, mannen går tillbaks till . back to work, the man returns
.. to . . .

High values for LV can result not only from effective lexical variation but also from an inability to construct a coherent text. If
the text is loosely composed topically, there are fewer opportunities to use the same content words many times. Similarly, high
values for LS can result from idiosyncratic and non-idiomatic use of the lexicon.
Thus, the quantitative measures employed in this study give only a gross indication of possible differences between the groups.
One would need to study the data qualitatively in order to get a more adequate picture, something we cannot do within the
present framework.
Lexical errors
As will presently be shown, the type and frequency of lexical deviances from the native norm were to some extent different in
the bilingual groups on the one hand and in the monolingual on the other. Initially, a description of the error categories with
examples from the data will be given.
As mentioned above, the two main categories of lexical deviances employed were approximations and contaminations. An
approximation is a deviant lexical unit that has such formal or semantic characteristics that it is immediately clear what the target
lexical element would be. The similarity between the unit produced and the target may be a matter of either form or meaning,
but it is often the case that both these categories coincide. A typical example of this is when the form njöts is used instead of
avnjöts as

< previous page page_75 next page >


< previous page page_76 next page >
Page 76
in lunchen njöts vid det rullande bandet (the lunch was taken (enjoyed) at the assembly line). The forms are passive past tense
forms of the formally and semantically similar verbs njuta and avnjuta respectively. Both mean 'enjoy' but the first one is
typically used as an intransitive and the second as a transitive verb. A contamination comprises erroneously combined elements
from two target forms.
The deviant unit may be either a single lexical unit or a lexical phrase. The lexical phrases are of various types, such as idioms,
other set phrases (where the content of the phrase is deducible from the content of its parts), and word constellations that are
determined by specific sub-categorisation and selection restrictions of Swedish.
The examples in 1-3 show approximations of a single lexical unit to a target form:
1. men hjälp av en bildskärm (Fi 5)
= med
(by means of a screen)
(men = 'but', med = 'by')
2. majskorv (Sp 14)
= majskolv
(corn-cob)
(-korv = 'sausage', -kolv = 'cob')
3. demostrera (Sw 32)
= demonstrera
(take part in a demonstration)
In 4-6 we find approximations of a single lexical unit to a target meaning:
4. den existerar i västra Sverige (Fi 6)
= förekommer
(it can be found in western Sweden)
(förekomma = 'can be found', existera = 'exist')
5. genom att ändra på det rullande bandets volym (Sp 16)
= hastighet
(by changing the speed of the assembly line)
(volym = 'volume', hastighet = 'speed'
6. på sommaren kan man se stora stim /av fjärilar/ (Sw 29)
= svärmar
(in the summer one can see large swarms /of butterflies/)
(stim = 'shoal', svärm = 'swarm')
Examples 7-9 are instances of approximations to lexical phrases.
7. den fungerar nog inte på längden (Fi 2)
= i längden

< previous page page_76 next page >


< previous page page_77 next page >
Page 77
(it probably will not work in the long run)
(på längden = 'lengthwise', i längden = 'in the long run')
8. hans uppmärksamhet riktas till en kraftig dam (Sp 14)
= mot
(his attention was turned to a fat lady)
(rikta uppmärksamhet mot is a set phrase)
9. då han alltid lyckas hamna i efterkälke (Sp 16)
= på efterkälken
(since he always succeeds in getting behindhand)
(på efterkälken is an idiom)
In the next examples, we find instances of contaminations. Examples 10-12 comprise contaminations of two single lexical units:
10. protestationståg (Sp 19)
= demonstrationståg
(demonstration procession)
(contamination of demonstration- = 'demonstration' and protest- =
'protest')
11. även den här hetsiska arbetsmomenten (Fi 4)
= hektiska
(even these hectic work components)
(contamination of hektisk = 'hectic' and hetsig = 'hasty')
12. mycket har blivit nerplånat (Sw 30)
= utplånat, nerrivet
(much has been wiped out-pulled down)
(contamination of utplånat = 'wiped out' and nerrivet = 'pulled
down')
The contaminations also occur in lexical phrases, as evident from Examples 13-15:
13. men i helhet tycker jag det var en jättetrevlig film
= i det stora hela
(but on the whole I found it to be a very nice film)
(contamination of i det stora hela = 'on the whole' and i sin helhet
= 'in its entirety')
14. börjar vrida runt majskorven i all hastighet (Sp 14)
= med hög hastighet
(starts to turn the corn-cob around with high speed)

< previous page page_77 next page >


< previous page page_78 next page >
Page 78
(contamination of med hög hastighet = 'with high speed' and i all
hast = 'in great haste')
15. man beräknar med att . . . (Fi 11)
= räknar med, beräknar
(one reckons that . . .)
contamination of räknar med = 'reckons' and beräknar -
'estimates')
The frequency and distribution of these error categories are displayed in Tables 2 and 3. The figures are given in numbers rather
than in proportions, since the total number of words produced by each group is very similar: 12,205 words in the Finnish group,
12,294 in the Spanish and 12,233 in the Swedish group.
Table 2. Frequency and distribution of approximations in written (W) and
oral (0) data
Words Phrases Total
W 0 W 0
Fi 21 (5) 12 (5) 11 15 59
Sp 22 (5) 12 (7) 24 12 70
Sw 10 (0) 5 (0) 4 2 21

Table 3. Frequency and distribution of contaminations in written


(W) and oral (0) data
Words Phrases Total
W O W O
Fi 2 1 6 3 12
Sp 1 2 6 3 12
Sw 0 5 3 6 14

As can be seen from the tables, there is a large difference between the groups with respect to number of approximations, while
the distribution and frequency of contaminations are practically the same in all groups. The number of contaminations, on the
whole, is very small. Thus, the more interesting results are those that concern the approximations.
In Table 2, we can see that the approximations are three times as frequent in the Finnish and three-and-a-half times as frequent
in the Spanish as in

< previous page page_78 next page >


< previous page page_79 next page >
Page 79
the Swedish group. Here, we thus have a highly significant difference between the bilingual students, on the one hand, and the
Swedish controls, on the other. The difference between the Finnish and the Swedish groups is the more striking as the Finnish
group generally had higher values on the quantitative measures presented above than the other two groups. And, as a matter of
fact, although all groups had the same average grade level in all school subjects, it turned out that the Finnish group had the
highest average grades in Swedish (mother tongue instruction) (3.7 against 3.1 for the Spanish and 2.8 for the Swedish group).
The figures in brackets in Table 2 indicate the number of approximations that happen to be other existing words in Swedish, for
example men (but) for med (by, with) or kvastar (brooms) instead of trastar (thrushes). These are fairly frequent in both the
bilingual groups, but do not occur at all in the Swedish group. This does not mean, of course, that such examples cannot be
found in monolinguals, but judging from this investigation, they may be more prevalent among near-native bilinguals than
among monolinguals.

Discussion
We have two topics to discuss against the background of our results, the differing proficiency level hypothesis and the notion of
fossilisation (cf. pp. 67-70). As will become evident, we shall have more to say about the second topic, since our data on the
first topic are difficult to interpret. At any rate, we shall initially focus on the topic of differing proficiency level.
If one were to assume that using a second language would generally hamper success in highly literate language tasks, one would
expect consistent differences between bilingual and monolingual subjects when confronted with the demands of such tasks. Such
a strong position may be disregarded, though, since it is not held by present day researchers (but it pops up in the general debate
about bilingualism in the school context), but also if the weaker claim were true, that a larger proportion among bilingual
speakers would risk not developing the proficiency level needed for literate language usealthough there were other bilingual
speakers that didone would similarly expect to find group level differences between monolingual and bilingual subjects.
The results of the present investigation do not support either of these positions. As we have noted, there are no consistent and
clear-cut differences between bilinguals and monolinguals on the measures employed to capture proficiency. The vocabulary, as
it is used in the literate tasks in this investigation, seems to be as large, as varied and as sophisticated in the bilingual groups as
in the monolingual group. Indeed, we do have individuals whose results indicate lower proficiency levels in vocabulary use, but
these individuals are distributed among all our subjects, both bilingual and monolingual.

< previous page page_79 next page >


< previous page page_80 next page >
Page 80
However, our study leaves a number of questions unanswered. One might want to explore, for example, what significance to
attach to the fact that the group of subjects studied in this investigation is special in certain ways. Most importantly, perhaps, our
subjects are chosen from the élite of their age group, i.e. they are students at the gymnasieskola (in the large majority of cases in
so-called theoretical tracks). What we do not know is whether bilingual and monolingual students who do not belong to this
group would exhibit the group level differences expected in the differing proficiency level hypothesis.
Secondly, even if we did find consistent differences between bilingual and monolingual students on the quantitative measures
we have employed, it would not be possible to immediately ascribe these differences to varying proficiency levels. It is well
known that cultural differences may manifest themselves in different ways of constructing texts (see Tannen, 1984; and Stroud,
in press), which might be an alternative explanation, if such results were obtained.
Thirdly, other-than-lexical measures of language proficiency might give results that are different from those that were
forthcoming here. However, in the larger investigation, of which this study is one part, the same subjects were also examined on
a number of syntactic variables that may relate to the requirements for literate language use. The picture drawn with respect to
those dimensions was similar to the one presented here (see Hyltenstam and Stroud, in preparation).
In summary, a more detailed qualitative analysis of the type of data we have might answer some of the questions put forward
and contribute to a more accurate picture of the language employed by bilingual and monolingual subjects in a highly literate
language task.
With respect to fossilisation, we have interpreted our results on lexical deviances as relevant to this notion, as the subjects
investigated can probably be considered to be in an 'end state' (Klein, 1986)if there is any substance to such notionsrather than in
a developmental phase in their Swedish. Our results on lexical errors show a striking difference between bilingual subjects on
the one hand and monolingual on the other.
The results are discussed in relation to the questions that were formulated above on fossilisation (cf. p. 68). Unfortunately, space
does not allow a treatment of all these questions. Therefore, we shall concentrate on points 4 and 5: Is fossilisation an age-
related or a pure second-language phenomenon (4) and, are there always some traces of fossilisation in second-language
acquisition contexts (5)? If fossilisation is an age-related phenomenon, second-language acquisition starting before puberty
would be characterised by the absence of fossilisation and identical in result to first language acquisition. If it is a pure second-
language phenomenon, features of fossilisation would be found in learners with second language onset both before and after
puberty.

< previous page page_80 next page >


< previous page page_81 next page >
Page 81
Our results indicate that fossilisation is not an age-related language learning phenomenon in the sense that it occurs only in adult
learners. All our bilingual subjects but one (who was 15) started their acquisition of Swedish before puberty. Actually, 12 were
below the age of five, six were 5-7 years old, and five were 9-12 when they came into contact with their second language. It is
important to note, however, that the frequency of errors is somewhat different in these three groups. In the written data from the
youngest onset group, the proportion of erroneous words was 0.6%, which means 2.6 words per subject. 3 In the next age group,
we have 0.9%, which is 4.1 words per subject, and among those who started late, the error proportion was 1.8% or 6.6 words
per subject. In the Swedish group, the proportion of error was 0.4%, i.e. 1.4 words per subject.
In the oral data, the results were not equally clear. The youngest onset group had an error frequency of 3.5 words per subject,
while the next youngest had 2.0 and the oldest 3.8. It is difficult to see why the middle age group, those who started at ages 5-9,
should have the lowest figures here. Interestingly, the Swedish group had a figure of 3.3 errors per subject in the oral data. This
means that the error frequency in the oral data is similar in all language acquisition onset groups, except for the middle bilingual
group. Given the small number of subjects in each age group, this result can very well have been obtained by chance, but if it
were duplicated in a larger study, one would need to go into a more detailed discussion of the optimum age hypothesis and the
effects of starting to acquire a second language at different ages (see McLaughlin, 1985, for an overview).
Looking at the difference between the bilingual and the monolingual groups in the written data, one might thus say that
fossilisation also occurs in children. Given the distribution of the errors, we have two possible interpretations. One is that since
there are more fossilised units in the learners who started their second language acquisition at a later age, age is still a relevant
factor, although the critical point may not be puberty. There may indeed be no critical point, but just a continuum, so that the
older the learner is by the onset of acquisition, the more fossilisation will make itself evident.
The second interpretation is that what we see in our data should not be characterised as fossilisation at all. It may well be that
the claim made lately, that it takes many years to achieve native-like command of a second language, is illustrated in our results.
In discussions about this question, it has generally been emphasised that children may acquire the formal aspects of a second
language quickly, but take up to 5-7 years to master the context-reduced cognitive language skills required for academic success
(again, cf. McLaughlin, 1985, for an overview). The results of the present study indicate that it may take many years to achieve
full command also of the formal aspects of a second languagealthough children acquire most of this within a short time span.
This observation can be related to another type of reasoning

< previous page page_81 next page >


< previous page page_82 next page >
Page 82
about the time it takes to acquire a second language, namely the discussion that focusses on the processing side of second
language abilities (cf. for example, Mägiste, 1986). Here it is suggested that processing capacity restrictions would account for
the fact that even bilinguals who are highly proficient in both their languages may make more errors and have longer response
times in certain psycholinguistic experiments than monolinguals. Apart from pointing to reasons why our bilingual students
make more errors than the Swedish subjects, perhaps also the difference between our written and oral data could be interpreted
in this framework. In writing, when there is more time for monitoring, the second-language processing limitations may not have
such an obvious effect. Therefore, learners who have spoken the second language for many years succeed in avoiding errors to a
larger extent here than in oral production. In oral production, furthermore, the lack of time for monitoring results in many errors
for native speakers of a language as well, which is clearly shown in our figures.
It is not possible, on the basis of this investigation, to select one of these alternative interpretations as the better one. One would
definitely need longitudinal studies of near-native speakers to see whether fossilisation features really are fossilised, or if they
disappear with time, although at a very slow rate.
In this section, we have discussed our two notions of interest, the differing proficiency level hypothesis and fossilisation, in
isolation from each other. One might consider whether there are any connections between them. Do features of fossilisation play
a part in hampering an efficient language use in literate contexts? If this were the case, the differences between our subjects on
error frequency would be relevant to our discussion of proficiency level differences. I think one could say that the lexical
deviances of the kind exhibited here to some extent, although limited, have a bearing on literate language proficiency. If the
meaning of lexical elements in a language is mistakenly represented by a speaker, this necessarily has consequences for
comprehension. In these data, however, many, or rather the majority, of the deviances from the norm concern formal aspects of
the lexicon and are not as such directly involved in the negotiation of meaning, which is central to literate language proficiency.
These formal deviances, obviously, play a minor role in the cognitive aspects of literate language use, since not only near-native
but also other advanced second-language speakers may be highly proficient linguistically in spite of fossilised elements in their
speech. One might suggest, instead, that the main effect of these deviances concerns the sociopsychological aspects of
communication.

Conclusion
In studying the use of lexicon among adolescent bilingual near-native and monolingual speakers of Swedish, we have found no
consistent differences

< previous page page_82 next page >


< previous page page_83 next page >
Page 83
between the groups on measures of lexical density, lexical variation and lexical sophistication. These measures were used to
investigate whether differences could be found between near-native and native speakers in highly literacy-related language
tasks. The aim was to provide a contribution to the empirical research on differing proficiency levels among near-native second-
language learners.
In studying lexical errors in the near-native and native speaker groups respectively, clear differences were obtained in frequency
of errors and also in the distribution of error types. The differences were discussed against the notion of fossilisation, and it was
concluded that fossilisation features occur not only among second-language learners who start their acquisition after puberty, as
seems to be the general assumption in discussions of this notion, but also among younger learners. With respect to the notion of
fossilisation, an alternative interpretation of the results was suggested, according to which what we see as fossilised features in
our population are not really fossilised but instances of error phenomena that disappear at an extremely low rate.

Notes
1. This study reports on work carried out in a project on the language of bilingual students in the Swedish gymnasieskola, who
are near-native second-language speakers of Swedish. The project was initiated and supported by the Swedish National Board of
Education. Financial support was also obtained from the Council of Södertälje and from the School of Education in Stockholm.
The team working on the project comprises Kenneth Hyltenstam and Christopher Stroud (principal investigators), Pirkko
Bergman, Kari Kuusisto, Alli Risberg and Lars Sjödin (research assistants).
Kari Kuusisto carried out the analyses of lexical density, lexical variation and lexical sophistication presented here. He also
undertook the statistical analyses of the data.
2. The formula is , where N is the number of tokens and V the number of types. k is a constant. (See Hultman and
Westman, 1977: 264).
3. The different number of words produced by each student has been considered by the calculation of these frequencies.

References
Biber, D. (1986) Spoken and written textual dimensions in English: Resolving the contradictory findings. Language 62, 384-
414.
Brown, G. (1981) Teaching the spoken language. In B. Sigurd and J. Svartvik (eds,, AILA 81, Proceedings II. Lectures. Special
issue of Studia Linguistica 35, 161-82.
Coppieters, R. (1986) Competence differences between native and fluent non-native speakers. Manuscript Pomona College.
Cummins, J. (1979) Cognitive academic language proficiency, linguistic interdependence, the optimum age question and some
other matters. Working Papers on Bilingualism, 19, 121-29.
(1983) Language proficiency and academic achievement. In J. W. Oller Jr (ed.) Issues in Language Testing Research. Rowley,
Mass.: Newbury House.
Edelsky, C., Altwerger, B., Barkin, F., Flores, B., Hudelson, S. and Jilbert, K. 1983)

< previous page page_83 next page >


< previous page page_84 next page >
Page 84
Semilingualism and language deficit. Applied Linguistics, 4, 23-41.
Hultman, T. and Westman, M. (1977) Gymnasistsvenska. Lund: Liber Läromedel.
Hyltenstam, K. and Stroud, C. (in preparation) Tvåspråkiga gymnasieelevers svenska. Stockholm University, Dept of
Linguistics.
Ioup, G. (1984) Is there a structural foreign accent? A comparison of syntactic and phonological errors in second language
acquisition. Language Learning, 34, 1-17.
Kellerman, E. (in press) The imperfect conditional: Fossilization, crosslinguistic influence and natural tendencies in a foreign
language setting. In K. Hyltenstam and L. K. Obler (eds), Bilingualism Across the Life Span. Cambridge: Cambridge University
Press.
Klein, W. (1986) Second Language Acquisition. Cambridge: Cambridge University Press.
Linnarud, M. (1975) Lexis in free production. Swedish-English Contrastive Studies, Vol. 6. Lund University: Dept of English.
(1986) Lexis in Composition. A Performance Analysis of Swedish Learners' Written English. Lund Studies in English, Vol. 74.
Lund: CWK Gleerup.
Mägiste, E. (1986) Selected issues in second and third language learning. In J. Vaid (ed.), Language Processing in Bilinguals:
Psycholinguistic and Neurolinguistic Perspectives. Hillsdale, New Jersey: Erlbaum.
Martin-Jones, M. and Romaine, S. (1986) Semilingualism: a half-baked theory of communicative competence. Applied
Linguistics, 7, 26-38.
McLaughlin, B. (1985) Second Language Acquisition in Childhood: Volume 2. School-Age Children. Hillsdale, New Jersey:
Erlbaum.
Selinker, L. (1972) Interlanguage. IRAL, 10, 209-31.
Stroud, C. (in press) Literacy in a second language: a study of text construction in near-native speakers of Swedish. Paper
presented at the 5th Nordic Conference of Bilingualism, Copenhagen, June, 1987.
Tannen, D. (1984) Spoken and written narrative in English and Greek. In D. Tannen (ed.), Coherence in Spoken and W'ritten
Discourse. Norwood, New Jersey: Ablex.
Ure, J. (1971) Lexical density and register differentiation. In G. E. Perren and J. L. M. Trim (eds), Applications of Linguistics.
Cambridge: Cambridge University Press.

< previous page page_84 next page >


< previous page page_85 next page >
Page 85

Syntax and Information Structure in Learner Language


Anne Holmen
The Royal Danish School of Educational Studies, 101 Emdrupvej,
DK-2400 Copenhagen NV, Denmark.
Abstract. The paper reports on the first part of a longitudinal study of young immigrants' spoken Danish during the first 3-
17 months of their stay in Denmark. The subjects include Albanian, Vietnamese, and English speaking adolescents who at
the time of recording attended schools in the Copenhagen area. The data was collected by means of audio-taped
interviews consisting of unguided conversation and elicited production (based on pictorial stimuli).
The study aims at comparing the developmental patterns of individual learners in order to investigate the role of their
linguistic and sociocultural background and present conditions and to gain insight into language learning processes. The
study focusses on syntactic development approached from a functional perspective, and in the part reported here, changes
in early second-language syntax are regarded as a result of changes in the ways in which information is organised in
learner utterances and thus rooted in cognitive as well as interactional dimensions of language use.

Introduction
Within interlanguage (IL) research one is concerned with second-language acquisition, i.e. with the conscious and subconscious
processes involved in learning a language other than the mother tongue whether through language instruction or in a natural
setting. Samples of learner data are studied in order to gain insight into the learning process and, if possible, to map out routes
taken by learners from their first acquaintance with a target language to a more advanced proficiency level. It is a common aim
of this line of research to be able to identify general and variable features of language learning by focussing on linguistic,
psychological and/or sociocultural aspects chosen according to the researcher's theoretical stand. Thus learners who differ, for
example, in age or mother tongue, in cultural and educational background, in attitude to the target language culture, or in
motivation for learning the target language are compared and the general validity of the differentiating factors discussed (Ellis
(1985) for an overview of current theories of second language acquisition).

< previous page page_85 next page >


< previous page page_86 next page >
Page 86
An important characteristic of learner language is continuous change. According to IL theory this change does not take place as
a mere approximation of mother tongue features to the target language norm, neither does the line of development seem to be
determined solely by the quality and quantity of the input to which the learner is exposed. Rather than seeing the learning
process as an interplay between two independent language systems (the mother tongue and the target language), the learner's
language is considered systematic in itself, i.e. it is governed by rules whichat any point in the learning processare internally
consistent (Tarone, 1982). According to this view, language learning takes place through the learner's activity in the second
languageby using it for communication and trying to make sense of input s/he is continuously engaged in forming, testing and
revising hypotheses about the target language (Færch et al., 1985) and thus gradually developing his her IL system. In these
processes an important role is played by the learner's previously obtained knowledge, i.e. all learners are expected to use the
strategy which McLaughlin (1978) calls 'employing what is known'. This may include linguistic knowledge of the mother tongue
and of the target language as well as knowledge of the world and personal experience with language learning.

Syntactic Development
Syntactic development has been the object of an extensive number of IL studies, a major part of which are concerned with the
nature of the IL continuum. The impact of transfer, complexification/simplification, and universal grammatical principles on the
learning process have been studiedmainly as (series of) static descriptions of IL forms (e.g. Dulay and Burt, 1974; Hyltenstam,
1978; Klein & Dittmar, 1979; Schumann, 1978). However, in order to understand the dynamic nature of IL change, we need to
know not only what syntactic structures are acquired, but also how and why this is so. In other words we need to know more
about the learner strategies involved in second language acquisition and about the interplay between cognitive and
social/communicative aspects (Fillmore, 1979; Kasper, this issue, p. 37).
One way of dealing with the 'why' of language learning seems to be in studying syntax through the changing patterns of
information structure (Huebner, 1983; Givón, 1979). Within this functional approach, cognitive and interactional dimensions
merge in so far as the basic question, how 'discourse uses syntax to promote understanding' (Hatch, 1983: 109), involves a broad
range of phenomena such as complexity and organisation of information, linguistic coding, context-relatedness, speaker
intentions and interactional roles.
In the case of early second-language acquisition, the functional approach seems even more relevant, if not the only approach
possible. In a very

< previous page page_86 next page >


< previous page page_87 next page >
Page 87
careful longitudinal analysis of one Hmong informant, Huebner (1983) reports that his original intention of studying the
changing tense-aspect system in the informant's English IL was abandoned as a preliminary analysis of the data clearly pointed
to the linguistic organisation of information as the most productive area of change.
In the early recordings, Huebner finds a preference for topic-comment structures expressed among other things through the
existence of a clear topic marker 'is(a)'. In later recordings, the form 'is(a)' expresses copula function and at the same time other
grammatical features such as subjects, definite articles and anaphoric pronouns appear.
The development depicted resembles the change from a presyntactic to a syntactic mode proposed by Givón for creolisation of
pidgins, diachronic change and first and second language learning (1979, 1984). Givón sees the two modes as poles in a
communicative continuum relevant to all languages and all communicative situations. In the pre-syntactic mode, word order is
mainly governed by pragmatic principles: old information goes first, new information follows, and topic-comment structures
occur. In the syntactic mode, word order is used predominantly to signal semantic case-functions andas a consequencesubject-
predicate structures prevail. Other changes along the continuum concern, e.g. the rate of delivery (slow?fast), the linking of bits
of information (through loose conjunction?tight subordination), and grammatical morphology (no?elaborate use).
According to Givón, all language development follows the same route along the continuum motivated by functional change in
the linguistic system. Thus the occurrence of topic-comment structures in early IL is based on the dominance of pragmatic
principles in word order. Schachter & Rutherford (1979), on the other hand, suggest that topic-comment structures appear in
some learners' IL as a result of structural or typological transfer from previously learned languages and, as a consequence,
learners take different paths when acquiring a given target language.
Hatch supports the idea of a common path of IL developmentat least for parts of the IL (e.g. negation). However, she does not
regard this as a result of natural changes in linguistic or cognitive terms, but rather as a reflection of the learner's growing
conversational competence: 'One learns how to do conversations, one learns how to interact verbally, and out of this interaction,
syntactic structures develop' (1978: 404). Later Hatch has slightly modified this view by suggesting that second language
acquisition may be jointly determined by factors inside and outside the learner: 'While social interaction may give the learner the
''best" data to work with, the brain in turn must work out a fitting and relevant model of that input' (1983: 180).
The present paper reports on the first part of a longitudinal study of young immigrants' interlanguage syntax through the early
stages of their acquisition of spoken Danish. The study aims at comparing the developmental patterns of individual learners in
order to investigate the role of their

< previous page page_87 next page >


< previous page page_88 next page >
Page 88
linguistic and sociocultural background and present conditions and to relate these findings to the different views on language
learning mentioned above. In this part of the study the development of early second-language syntax is regarded as a result of
changes in the ways in which information is organised in learner utterances.

Data
Data were collected from 10 young learners of Danish through the first 12-15 months of their stay in Denmark. On arrival they
were 15-17 years old and all attended Danish schools during the period of data collection. The six learners which were selected
from this group are shown in Table 1. There were two from each language background (Albanian (Gheg), Vietnamese and
English). So far only data from learners C and E have been analysed carefully, and the following presentation will therefore
focus on these two learners.
The data were collected by means of audio-taped interviews, each consisting of unguided conversation and elicited production
(based on pictorial stimuli). Each interview was audio-taped in its full length (20-40 minutes), and then transcribed in ordinary
orthography. The learners were encouraged to introduce topics for conversation or elaborate on the pictures, which were chosen
with the specific purpose of engaging adolescents in conversation. All produced a couple of spontaneous utterances in the first
interview session but seemed to prefer the picture description. Later sessions, however, include an increasing number of
narratives, personal comments, etc. and the picture description is still carried out, but not elaborated on.
On two occasions data of a slightly more formal nature were collected as the learners were asked to retell a story depicted in a
series of drawings a genre familiar to all of them from the Danish school. Afterwards they were asked a fixed set of questions
concerning the pictures. The two comparable events took place with seven months in between for the Vietnamese and
Table 1
Learner Nationality Sex L1 No. of interviews
B Yugoslav Male Albanian 5
C Yugoslav Male Albanian 7
D Vietnamese Female Vietnamese 7
E Vietnamese Female Vietnamese 8
G American Female English 4
H American Female English 4

< previous page page_88 next page >


< previous page page_89 next page >
Page 89
Albanian-speaking adolescents, but unfortunately with only three months in between for the English-speaking. The difference in
time span makes a comparison of the learners on this basis very uncertain. However, it is worth mentioning that for all learners
there was an increase in utterance length (measured in number of words); for the Albanian-speaking boys there was a further
increase in the number of well-formed sentences (in existence and order of obligatory constituents) and for the English-speaking
as well as the Vietnamese-speaking girls an increase in number of subordinate clauses and in the number of sentences with an
adverb in initial position. It seems likely that this general developmental trend reflects an increase in the amount of information
expressed by the learners, a tendency similar to one reported in studies of children's acquisition of Danish as a mother tongue
(Hansen, 1975).

Changes in the Information Structure


In the following pages, it will be assumed that two functional distinctions are basic for communication to succeed in any
language. One concerns given and new information in the utterances
given (or old) information is that knowledge which the speaker assumes to be in the consciousness of the addressee at the
time of the utterance. So-called new information is what the speaker assumes he is introducing into the addressee's
consciousness by what he says. (Chafe, 1976: 30).
The speaker must find linguistic ways of indicating what is old and what is new information in his utterances and thus bring
about referent identification and textual coherence while expressing non-redundant information (Givon, 1984).
At the same time, the speaker must be able to mark case relations of the words used so that, for example, agent and object (in
relation to the action) may be clearly identified.
The terms topic and comment are often used interchangeably with given/ new information. However, in this paper topic refers to
the setting or 'the frame in which the sentence holds' (Chafe, 1976: 51) and comment to the focus of the utterance.
Based on these distinctions it is possible to divide the IL development of learners C and E into four stages characterised in the
following way and with time indications in brackets (number of months after starting the Danish school):
I. only one piece of new information per utterance: invariable sentence structure (3-5);
II. time and place indications added: variable sentence structure (5-7);
III. breaking up complex information into a line of 'det er' utterances (8-10);
IV. embedding (15-17).

< previous page page_89 next page >


< previous page page_90 next page >
Page 90
The time indications refer to the time of recording (and do not necessarily give a true picture of the learners' rate of learning). As
there are no clear boundaries between the stages, they must be read with caution.
I
Only one piece of new information per utterance: invariable sentence structure
During this stage the two learners either introduce new referents by using the phrase 'det er NP' (it is NP) or apply new
information to a referent given by me in the previous utterance (Example 1-3):
(The examples are given in Danish orthography in the left hand column with an English word-by-word translation in the right
hand columnwhen necessary an interpretation is added in brackets; E is the Vietnamese girl, C the Yugoslav boy and A the
present author.)
1 A: hvad kan du se der? what can you see there?
E: det er en havn it is a harbour
2 A: hvad gør den mand? what does that man?
E: iden mand øh mand han han that man er he he points at watch
maybe
peger til ur måske
3 A: det er en fransk bil it is a French car
C: fransk ja koster mange penge French yes costs much money

Only on a few occasions do the learners need to introduce a referent and at the same time apply a comment. This is done by
combining the two structures (in Example 4 it is important to know that E's brother had not been mentioned previously) or by
making a loose concatenation (Example 5):
4 E: det er min lillebror han spiller it is my little brother he plays
trommer drums
5 A: hvad er det? what is this?
C: det er en dreng og en it is a boy and a motorbike
motorcykel

Almost all utterances are simple (i.e. containing only one verb). In the case of complex utterances, the verbs are linked by being
listed in order of actual occurrence (Example 6):
6 E: den åbner kommer vand luk it opens comes water close
(about a tap)

In addition both learners manage to contribute to the dialogue through

< previous page page_90 next page >


< previous page page_91 next page >
Page 91
repetitions of words or phrases said by me, often with a yes or no preposed (Examples 7-8):
7 A: kører hun med toget? goes she by train?
E: ja hun med toget og bus yes she by train and bus
8 A: om vinteren skal man have trøjer in winter must one wear
på i Vietnam? sweaters in Vietnam?
E: ja trøjer ja Vietnamikke sne der yes sweaters yes Vietnamnot
snow there

Learner E produces only a few sentences which resemble topic-comment structures (Example 9) i.e. in the beginning of the
utterance the frame is given, followed by the comment, which is also new information, without any syntactic ties. C produces no
such structures.
9 A: hvad bruger du den til? what use you that for?
E: den der spise morgenmadeller that one eat breakfastor drink
drikke kaffe sukker coffee sugar

In general both learners manage to communicate in very short utterances which all follow the principle of one piece of new
information in each utterance. The new information is typically an agent or verb, but never a verb followed by its complement,
which may account for the fact that there are no instances of confused agent/object relations. The need to introduce new
referents (as agents) and name actions performed by agents already mentioned seems to bring about a limited set of utterance
structures. The new information may, of course, also consist of an affirmation or negation of (part of) the question asked, which
is then repeated in the learner's utterance as old information (Examples 7-8).
II
Time and place indications: variable sentence structure
During this stage more bits of information are introduced into the utterances, mainly through indications of time and place
(Examples 10-14). There are only few attempts at subordinating clauses (Example 14).
10 C: på Jugoslavien jeg In Yugoslavia I understand paper
forstår avis
Dan' jeg ikke forstår Denm' I not understand
11 C: jeg laver på I make at Nørrebro stadium it has many
Nørrebrohallen det har boys at sport (I do sport at N. there are
mange drenge på sport many boys doing sport)

< previous page page_91 next page >


< previous page page_92 next page >
Page 92
every day evening my father he
12 E: hver dag aften min far han reads Danish I and NAME she
læser danske jeg og NAVN hun writes Danish
skriver danske
maybe NAME she speaks I
13 E: måske NAVN hun taler jeg understand very well
forstår meget godt
Peter he says as he looks he looks
14 E: Peter han siger da han kigger at it Peter says he says stop it
han kigger på den Peter siger han
siger stop den

Examples 15-16 are the only recorded attempts at expressing a causal relation produced by the two learners during the first
seven months of their acquisition of Danish:
he is fat he eats (later)
15 E: han er tyk han spiser (senere) because he eats
fordi han spiser
16 C: han er tyk spiser he is fat eats

During this stage the first instances of the learners' openly correcting their own word order appearmaybe this is an indication of
the first awareness of the grammatical function of Danish word order (Examples 17-18).
17 A: Wuzka er det et navn? Wuzka is that a name?
Wuzka in Yugoslavia
C: Wuzka på Jugoslavien
yes
A: ja
stupid (in)
C: dum i
means it stupid?
A: betyder det dum?
no stupid stupid is called
C: nej dum dum hedder
stupid
A: dum
yes
C: ja
yes is it him
A: ja er det ham der
yes is called stupid
C: ja hedder dum
it is maybe king he likes
18 E: det er måske konge han kan lide dogor dog loves king
hundeller hund elsker konge

During Stage II learner E to a very large extent keeps the basic word order pattern established above, i.e. (adv) SV (O) (adv) or
'det er' + NP. Learner C, however, changes between SV and VS word order with intransitive verbs and often leaves out the
anaphoric pronoun (compare Examples 19 and 20, and 17 above):

< previous page page_92 next page >


< previous page page_93 next page >
Page 93
19 A: er du sammen med NAVN? are you together with NAME?
C: ikke kommer NAVN not comes NAME
come I like thiscar (pointing)
20 C: kommer jeg sådanbil (peger) comes here
kommer her

III
Breaking-up complex information into a line of 'det er' utterances
After Stage II, in which more information was added to the nucleus, one might expect attempts at integrating these bits of
information into the utterances by embedding. However, for a couple of months the two learners seem to choose a different
strategy, namely that of speaking in fairly long sequences which put together contain quite a number of different pieces of
information, but which are not integrated into one syntactical unit. The phrase det er (it is) is now used for another purpose
besides introducing new referentsthat of linking bits of information placed in short parallel sentences:
it is Saturday it is every week it is my
21 E: det er lørdag det er hver grandmother my aunt they come and
uge det er min mormor min visit my family
moster de kommer og besøger
min familie
it is more more learn (in this way we
22 E: det er mere mere lærer learn more)
it is guests it is dry the hand dry the
23 E: det er gæster den er mouth and it is much must I make ten
tørrer hånden tørrer munden og (she is supposed to sew ten napkins for
det er meget skal jeg lave ti a wedding)
it is operation it is not finished with
24 C: det er operation det er that
ikke færdig med den
it is a pilot it is stop plane in
25 C: det er en pilot det er stop aerodrome (airport)
maskine på aerodrom

IV
Embedding
During the last interview, which takes place approximately six months after the previous one, the pattern of breaking up
information into short, parallel sentences has been replaced by attempts at making subordinate clauses or infinitive constructions,
thus integrating the more complex information into syntactical units (Examples 26-27). Neither of the two learners seems to
experience any serious problems with word order in the main

< previous page page_93 next page >


< previous page page_94 next page >
Page 94
clause. In the subordinate clauses, however, both display a tendency to leave out anaphorical pronouns, (Example 28) and in
main clauses following preposed subordinate clauses, there is variation between SV and VS order:
Danish friends no I have not so
26 C: danske venner nej jeg har much now because I have not time
ikke så meget nu fordi jeg har ikke to go out . . . so I have not so
tid gå i byen . . . så jeg har ikke så much time with the girls to go out
meget tid med pigerne at gå i byen
because some of the children they
27 E: fordi nogle af børnene de de they expect her mother er her
venter hendes mor øh hendes mor mother be pick them up
være henter de hjem
NAME she knows not how so stays
28 E: NAVN hun ved ikke hvordan home the (teacher does not know
så bliver hjemme how to make the girl come to
school so the girl stays home)

To summarise, the similarity between the interlanguage development of learners C and E when described in terms of
information structure has brought about the setting up of four stages of development. It is thus suggested that the ordering of
elements in the utterance, i.e. syntax proper, is secondary to the need to organise information in the utterances. Once the basics
of the discourse level are established, the learner apparently starts reorganising his utterances according to rules of grammar
instead of rules of information structure.
It seems likely that the learner's IL is more open to typological transfer from the mother tongue during the syntactisation process
as the learner's hypotheses about the target language focus on syntactic questions. This may explain a difference between the
two learners only hinted at previously: when learner E after Stage I attempts to speak Danish, she adheres to a fairly rigid word
order pattern, whereas C displays variable word order to a larger extent.
As an answer to the question 'what do the two learners achieve in their IL at different periods during the first year of
acquisition?' the four stages may be characterised in the following way:
During Stage I the learners are concerned with surviving communicatively. They draw on all the set phrases that they know of,
repeat as many of their interlocutor's words as possible and limit their creative use of the language to the principle of one piece
of information at a time. When this is not sufficient and when the communicative pressure gets too hard, they may produce
topic-comment utterances.
During Stage II a bit more information is introduced into the utterances, mainly modality expressions of time and place.
Instances of variable syntax may indicate that an awareness of grammatical functions of word order is

< previous page page_94 next page >


< previous page page_95 next page >
Page 95
on its way, but this is still not properly rule-based and there are many examples of concatenated pieces of information.
Stage III is perhaps the actual starting point of the syntactisation process. The frequent breaking up of the informational pattern
into smaller parallel units seems to reflect a need to express more complex information, yet the mere concatenation of words
found in the early stages is not sufficient, as this does not make marking of grammatical relations between the words possible.
At Stage IV the process of integrating more complex structurescontaining more than one verb and/or expressing temporal,
causal and conditional relationshas started.
Negation
The study of negation in the two learners' Danish sentences lends support to the division of their IL development into the stages
mentioned (Bolander, this journal, p. 97):
During Stage I, negation is produced exclusively by an anaphoric 'no' or by external negation:
29 A: er det en gammel is it an old man?
mand?
E: nej en gammel no an old
30 A: kender du en dansk know you a Danish girl?
pige?
C: nej jeg har no I have (no I don't have a Danish girl-
friend)

During Stage II external negation still occurs but there are also instances of internal negation, placed before main verbs:
it is NAME decides I not
31 C: det er NAVN bestemmer jeg ikke decide
bestemmer
32 E: unge i Vietnam ikke drikker øl young people in Vietnam not
drink beer

At the start of Stage III there is no more external negation, but exclusively pre-verbal replacement of the negative word 'ikke'.
As this stage includes only one interview with learner C, I have no way of seeing the next development in his IL before Stage
IV. Learner E, however, proceeds from a phase of pre-verbal negation to a distinction between auxiliary verbs and main verbs:
I have not read (English) in
33 E: jeg har ikke læse (engelsk) i Vietnamnow they not learn
Vietnamnu de ikke lærer engelsk English

< previous page page_95 next page >


< previous page page_96 next page >
Page 96
At Stage IV both learners consistently place the negation post-verbally in main clauses as well as in subordinate clauses.
The use of internal negation appears during Stage III, which was previously suggested as a starting point for the syntactisation
of the IL.

Conclusion
Clearly, one may question the general validity of the four stages set up on the basis of limited data collected from only two
learners. However, the qualitative analysis of these data has generated some hypotheses about the nature of early second-
language learning, which will be tested on the full group of learners in the next part of the study. The hypotheses are the
following:
1. early acquisition of syntax is determined by pragmatic principles;
2. the pragmatic period is followed by a period of syntactisation; and
3. typological transfer occurs as part of the syntactisation process.

References
Chafe, W. L. (1976) Givenness, contrastiveness, definiteness, subjects, topics, and point of view. In C. N. Li (ed.), Subject and
Topic. New York: Academic Press.
Dulay, H. and Burt, M. (1974) Natural sequences in child second language acquisition. Language Learning, 24, 37-53.
Ellis, R. (1985) Understanding Second Language Acquisition. Oxford: Oxford University Press.
Færch, C., Haastrup, K. and Phillipson, R. (1985) Learner Language and Language Learning. København: Gyldendal.
Fillmore, L. Wong (1979) Individual differences in second language acquisition. In Fillmore et al. (eds), Individual Differences
in Language Ability and Language Behaviour, New York: Academic Press.
Givón, T. (1979) From discourse to syntax. Grammar as a processing strategy. In T. Givón (ed.), Discourse and Syntax. New
York: Academic Press.
(1984) Syntax. .A Functional-Typological Introduction. Amsterdam: Benjamins.
Hansen, E. (1975) Syntaksen i børnesprog. Faglig-pædagogiske små-skrifter om dansk sprog. 4. København: Gyldendal.
Hatch, E. (1978) Second Language Acquisition: A Book of Readings. Rowley, Mass.: Newbury House.
(1983)Psycholinguistics: A Second Language Perspective. Rowley, Mass.: Newbury House.
Huebner, T. (1983) A Longitudinal Analysis. The Acquisition of English. Ann Arbor: Karoma Publishers.
Hlyltenstam, K. (1978) Variation in Interlanguage, Syntax Working Papers No. 8. Lund: Dept of General Linguistics, Lund
University.
Klein, W. and Dittmar, N. (1979) Developing Grammars. Berlin: Springer.
McLaughlin, B. (1978) Second Language Acquisition in Childhood. Hillsdale, NJ: Erlbaum.
Schachter, J. and Rutherford, W. (1979) Discourse function and language transfer. Working Papers on Bilingualism, 19, 1-12.
Schumann, J. (1978) The Pidginization Process. Rowley, Mass.: Newbury House.
Tarone, E. (1982) Systematicity and attention in interlanguage. Language Learning, 32, 69-82.

< previous page page_96 next page >


< previous page page_97 next page >
Page 97

Is There Any Order? On Word Order in Swedish Learner Language


Maria Bolander
Department of Scandinavian Languages,
University of Umeå,
S-901 87 Umeå, Sweden.
Abstract. This paper reports on a study of the acquisition of rules for the placement of the negative particle and some
adverbs (mostly sentence adverbials) in Swedish. These elements may have the same position in main as well as in
subordinate clauses. The study is based on three types of data from 60 tutored adult learners of Swedish with Finnish,
Polish and Spanish as their native languages.
In part, the results support earlier studies on favourable contexts for the correct application of the placement rules. Yet, an
interesting outcome of this study is the learners' placement of negation post-verbal to non-finite verbs in verb groups. An
error analysis shows tendencies for negation of constituents and early phrasal learning of frequent and useful sequences to
provide constraints on the application of the syntactic rules later. In addition, in many cases the errors seem to be caused
by difficulties in the retrieval of adequate lexical units. This may lead to violations in the application of syntactic rules
correctly applied in other contexts.
Finally the results indicate influence of input by later learning of rules with variable input. This is the case for adverbs,
despite Swedish grammar teaching focussing on them as following the same rules, and the placement of negation in some
subordinate clauses with variable placement of negation.

Introduction
A large number of studies have dealt with different aspects of the universal category negation (e.g. Horn, 1978; Dahl, 1979;
Sennekampf, 1979; Payne, 1985). Givón (1984: 322), for example, states that 'in naturallanguage unlike formal logicnegation is
a complex functional domain, drawing on three distinct though partially interdependent components', namely propositional
semantics, subjective certainty and discourse-pragmatics.
During the last two decades, research on the acquisition of negation in first as well as in second language acquisition (e.g. Klima
& Bellugi, 1966;

< previous page page_97 next page >


< previous page page_98 next page >
Page 98
Cazden et al., 1975; Wode, 1981; Hyltenstam, 1978; Felix, 1978; Ravem, 1978; Clahsen, 1984) has suggested certain
hypotheses on the developmental order of negation for English, German and Swedish. Pre-verbal placement seems to be an
early 'stage' in all acquisition of negation (NEG), and auxiliaries (AUX) are the favoured context for placement of NEG when
compared to mV (modal and main verb) and V (main verb) contexts.
The forms of standard negation may vary in different languages. However, typological studies have shown a universal
preference for a pre-verbal position (Dahl, 1979). For this study all present native languages of the subjects have the more
unmarked pre-verbal position of negation in contradistinction to the case for Swedish main clauses. In Swedish the negative
particle inte (NEG) and some adverbs (ADV) have the following positions:
1. MAIN CLAUSES
i. SVfinADV/NEG0
Han läser nog/ofta/inte böcker
(He reads probably/often/not books)
ii. SAuxfinADV/NEGVnf0
Han kan nog/ofta/inte läsa böcker
(He can probably/often/not read books)
2. SUBORDINATE CLAUSES
i. COMPSADV/NEGVfinO
när han nog/ofta/inte läser böcker
(when he probably/often/not reads books)
ii. COMPSADV/NEGAuxfinVnf0
när han nog/ofta/inte kan läsa böcker
(when he probably/often/not can read books)
As shown above the positions for NEG also hold for sentence adverbials and some other constituents (other adverbials and
some predicatives). A question here is whether the learners perceive the similarities with respect to the placement of NEG and
ADV, which is often emphasised in teaching, or if they have different rules for the placement of these two categories. One
reason for assuming that these two elements may be assigned positions by different rules is that Swedish also allows an initial
and/or final placement of ADV in most cases. Thus, the input concerning the placement of ADV is more variable and may
therefore confuse the learners (cf. Bourgonje, 1984).
A further complication concerning the placement of NEG and ADV is that some Swedish subordinate clauses often have main
clause word order, e.g. att (that) clauses. This is the case when such clauses contain asserted information (Lindberg, 1973;
Andersson, 1975). This fact is often ignored in teaching with the consequence that the learner has to handle the variable input
without knowing the underlying rules.
The present study attempts to gain further knowledge of favourable linguistic contexts for the placement of NEG according to
the Swedish norm

< previous page page_98 next page >


< previous page page_99 next page >
Page 99
in spontaneous speech in main and subordinate clauses, i.e. subordinate clauses with invariable as well as variable placement of
NEG. Another purpose is to compare the subjects' placement of NEG to their placement of ADV with the same positions in
order to investigate the effects of the native languages as well as the effects of the variable input compared with focus of the
instruction on the similarities between the placement of NEG and ADV.
Furthermore, the learners' interlanguage and their processing is studied through investigation of errors in ADV and NEG
placement. Finally, how the learners apply the rules in three different situational contexts is briefly discussed.

Methods
The study is based on three types of data from 60 adult learners of Swedish as a second language. The data can be said to be
longitudinal since they were gathered during a period of four months during which the learners were taught Swedish at a Labour
Training Centre. The learners are 20 Finnish, 20 Polish and 20 Spanish native speakers, and within each group there are 10 low
level learners (named F, P or S, plus a number between 01-26) and 10 high level learners (with the numbers 51-76). The
relatively large number of subjects, the three different native languages and the two levels of proficiency allow us to use the data
for cross-sectional studies and the groups are shortened FH for Finnish high-level learners, FL for Finnish low level and so on.
The three types of data consist of recordings of speech from (1) a ten-minute interview on selected topics and (2) descriptions of
a set of pictures. In addition, the taped data were complemented with (3) an acceptability test comprising written sentences (see
below). The recordings were made on three occasions: at the beginning, in the middle and at the end of the Swedish course. The
data used for this study are from the first and third occasion. Each recording session will henceforth be abbreviated as Time 1
and Time 3; the different data types are referred to as II, and 13, Pil, and Pi3 for the interviews and picture descriptions
respectively. (The second interview is excluded since there were only small differences in the categories studied between the
three occasions of recordings.)
All complete and comprehensible clauses containing NEG and ADV have been allocated to different categories according to the
verbal context of NEG and ADV and grammatical correctness with regard to the target language norm. Unfortunately, there is
no reasonable way to characterise the learners' interlanguage except in terms of deviations from the target language norm.
Hence, the inappropriate concept of 'grammatical correctness' is still used in this report. 'Formula' clauses containing X vet inte
(X doesn't know) are excluded (cf. Cazden et al., 1975; Clahsen et al., 1983).

< previous page page_99 next page >


< previous page page_100 next page >
Page 100
Due to the variable placement of NEG and ADV in certain Swedish subordinate clauses in native speakers' language, the
subordinate clauses are divided into 'subordinate clauses' and 'att' (that) clauses. The latter group, which may have either main
or subordinate placement of NEG and ADV depending on the information structure of the clause, consists of clauses introduced
by the subordinators att (that), därför (att) (because) and så (att)/ så . . . att (so that).
The data are displayed both as relative frequencies and absolute numbers in order not to be misleading, since the numbers are
sometimes small. An implicational scale is presented for subjects who produce errors in main clauses containing NEG. Due to
the small numbers and mostly similar tendencies for the learners within the six different groups, the results in this paper are
largely given as 'group results', which are complemented by an error analysis. The methods, all the data (including a large
amount of raw data) and the results are comprehensively accounted for in Bolander (1987).
This paper will first present the results of the negation study, including a context and an error analysis of the placements of NEG
in main clauses as well as in the different types of subordinate clauses. In the following sections there will be a comparison
between the placement of NEG and the placement of ADV in main clauses, and a short comment on the three different data
types before some concluding remarks.

Results and Discussion


Negationan overview
With regard to the total use of NEG, the Polish and the Spanish high level learners (PH and SH) use more of this category than
the other groups in main (NEG-mains) as well as in subordinate clauses (NEG-subs) (see Table 1). All learner groups, including
the high level learners who already seem to have fairly good knowledge of Swedish at Time 1, increase their use of NEG from
Time 1 to Time 3, though this does not hold for every single subject.
The increase in use of NEG is mostly found in holophrases (e.g. jag vet inte (I don't know) and the negative quantifier ingen
(nobody) and its flexion forms, as well as negation of single constituents outside full clauses. Thus, the increase of NEG in full
clauses at Time 3 does not seem to be a result solely of avoidance of NEG in full clauses at Time 1 but an indication that 'the
complex functional domain' of negation (Givón, 1984) makes NEG a phenomenon that requires more than language
development to be realised in real communication. However, for instance, the number of NEG-mains on one and the same
occasion varies between the subjects from 4 to 31 and without a deep analysis it is difficult to explain what constitutes a
communication containing many NEGs.

< previous page page_100 next page >


< previous page page_101 next page >
Page 101
Table 1. Frequencies for NEG in different contexts in main and subordinate clauses (upper values,
numbers of correct/total; lower values, relative frequencies)
V COP M mV PERF VV Tota
Learners Main Sub Main Sub Main Sub Main Sub Main Sub Main Sub Main Sub
FL 122/139 0/8 43/43 0/2 12/12 0/1 52/54 0/4 9/9 0/1 2/2 240/259 0/16
0.88 0 1 0 1 0 0.96 0 1 0 (1) 0.93 0
FH 107/110 2/8 72/72 2/2 18/18 0/4 30/30 26/28 1/1 10/10 0/1 263/268 5/16
0.97 0.25 1 (1) 1 0 1 0.93 (1) 1 0 0.98 0.31
PL 141/148 1/10 48/48 1/7 30/30 40/40 0/3 1/1 4/5 264/272 2/20
0.95 0.1 1 0.14 1 1 0 (1) 0.8 0.97 0.10
PH 118/12013/41 89/89 1/6 11/12 1/1 80/80 1/6 19/21 0/1 5/5 0/2 322/327 16/57
0.98 0.32 1 0.17 0.92 (1) 1 0.17 0.90 0 1 0 0.98 0.28
SL 96/114 1/12 49/49 27/28 45/52 1/1 5/5 223/249 1/12
0.84 0.08 1 0.96 0.87 1 1 0.90 0.08
SH 149/150 5/16 63/63 2/3 25/25 0/3 93/94 4/9 12/13 3/6 9/9 351/354 14/37
0.99 0.31 1 (0.33) 1 0 0.99 0.44 .92 0.5 1 .99 0.38
Total 733/78122/95 364/364 6/20 123/125 1/9 340/350 5/22 68/73 4/9 35/36 0/3 1663/1729 38/158
0.94 0.23 1 0.30 0.98 0.11 0.97 0.23 0.92 0.44 0.97 0 0.96 0.24
Abbreviations: main, main clauses; sub, subordinate clauses; V, simple main verbs; COP, copula; M,
simple modals; mV, modal + main verb; PERF, perfect or past perfect tense; VV, other verb groups.

< previous page page_101 next page >


< previous page page_102 next page >
Page 102
The results show that 30 of the subjects have acquired the rule for the placement of NEG in main clauses before the onset of
this investigation, and that most of the others make only a very few errors. Twelve of the learners make errors at Time 1 only,
thus exhibiting development of the rule, but 14 make errors only at Time 3 and five on both occasions. Frequency of negation in
subordinate clauses is much lower and the occurrence of errors substantially larger.
Thus, as a whole, there is no apparent development in all these learners' application of the studied rules from Time 1 to Time 3,
but for some of the least-developed learners there are more sentence negations at Time 3.
NEG in main clausesanalysis of linguistic contexts
In order to test whether the results reported in Hyltenstam's study (1978) on elicited data also apply to this study of spontaneous
speech, the subjects' placement of NEG in the context of different types of verbs in main clauses was investigated. The contexts
considered were: simple main verb (V), simple modal verb (M), the copula (COP) and AUX+main verb (vV). In a more detailed
analysis, vV is divided into modal verb+main verb (mV), perfect past perfect tenses (PERF) and other verb groups (VV) (see
Table 1).
The different contexts are displayed on the implicational scale in Table 2, which displays only those subjects who make errors
in NEG-mains. The results resemble those of earlier studies in the sense that most errors are made in main verb contexts (V)
and none or few in the contexts of COP and M. This pattern is found for all subjects but one, and the similarities are very likely
the result of an interaction between first language interference and universal factors.
Well worth noting in Table 2 is that simple modals would seem to favour a position of NEG according to the Swedish norm, as
seen by the high percentage of correct NEG placement. This is not unexpected but, in fact, this has never been stated for
Swedish before, since Hyltenstam's study (1978) only consisted of mV, which gave scope for the interpretation that the correct
placement of NEG in mV was best stated as a placement before the main verb instead of after the modal (cf. Hammarberg,
1979). However, the results from this study do not lend credence to such a hypothesis (see further below).
In order to, at least partly, reveal the interlanguage of the subjects, an error analysis may be of interest in order to investigate
the form-function relationships and possible keys to interlanguage processing.
Out of 48 errors in the context of V at least 24 of them may be interpreted as negation of single constituents. The basis for such a
hypothesis is that NEG in all these clauses is placed immediately before a constituent that is opposed to some element in the
near context, for example, the clauses are preposed by men (but), or there are two constituents directly contrasted to

< previous page page_102 next page >


< previous page page_103 next page >
Page 103
Table 2. The NEG placement rule in main clauses for learners
producing one or more deviations from the rule (+, means
according to the rule;, deviant from the rule and signs with ( )
are representing just one occurrence)
Learner COP M vV V
F03 + (+) ()
F07 + + () +
F09 0 0 +
F17 + (+) 0 ()
F22 + + +
F26 + (+) +
F55 + 0 + ()
F56 + + +
F64 + + () +
F66 + (+) () +
P01 + 0 () +
P03 + + +
P06 + + + ()
P16 + + + ()
P18 + 0 0 ()
P20 + 0 (+)
P51 + () + +
P52 + 0 () +
P64 + 0 + ()
P53 + + () +
S01 + + () ()
S05 + (+) +
S09 + + + ()
S10 (+) + ()
S11 (+) + +
S13 + () (+)
S15 + (+) +
S16 + + ()
S54 + + () ()
S55 + 0 () +
Relative frequency 1 0.92 0.56 0.35
Abbreviations: M, simple modal verb; COP, copula; vV,
AUX+main verb; V, simple main verb.

< previous page page_103 next page >


< previous page page_104 next page >
Page 104
each otherflicka gav och inte tar mat (girl gave and not takes food). Clahsen et al. (1983) have propounded that negation in all
early learner language is negation of constituents. Perhaps this is true also for more developed learner language, when a learner
wants to negate a special constituent and is ignorant of the accurate way of doing so in the target language.
With regard to the subgroups of the context vV, the absolute numbers and the relative frequencies for each context are presented
in Table 1 (see above). Totally, by far the most common context of the subgroups mV, PERF and VV is mV (340/350), which
contains the least errors relatively. The same value for relative frequency holds for VV, but for this subgroup the number of
occurrences is no more than 36.
In 3 out of 17 error clauses containing vV, NEG is placed before the finite verb. However, two of these clauses have NEG
placed sentence-initially and the third contains the verb group brukar använda (uses to use)a finite verb somewhat different
from the 'ordinary' auxiliaries. Thus, in all the other error clauses NEG is placed after the non-finite verb.
Placement after the non-finite verb is a matter that has been discussed very little in previous works. In Hyltenstam's study
(1978) based on elicited data, the only possible placements were either immediately before or immediately after the finite verb,
e.g. hankansluta (hecanstop). For spontaneous speech, however, the position after the non-finite seems to be a real alternative
for the learners of Swedish, indicating that different types of data may render somewhat different results.
In all but one of the NEG-mains containing NEG after the non-finite verb another element may follow after the non-finite verb,
for example
1. men vi ha fått inte (F64:I1)
(but we have become not)
2. ja ska prata inte svenska (F07:I1)
(I will speak not Swedish)
3. han kan säja inte så mycke (S05:I1)
(he can say not very much)
One possible explanation for such errors could be that the learner has noticed that NEG is to be placed after the verb, but that
s/he operates the verb group (mV/PERF) as a unit, some kind of prefabricated pattern (cf. Hatch, 1973; Hakuta, 1974). Perhaps
such a hypothesis is not too preposterous for these tutored learners who are certainly trained in tense flexion. An additional
support for this hypothesis may be the following adjustment caused by feedback from the speaker herself:
4. ja har läst inte ja har inte läst (P10:I1)
(I have read not I have not read)
Of course, it is not possible to know the way of the learner's processing, but a speculation might be that the perfect tense in the
more frequent

< previous page page_104 next page >


< previous page page_105 next page >
Page 105
affirmative form har läst during the speech production emerges as a unit and NEG is postponed, until she hears what she is
saying. Thus, the adjustment reveals that the learner knows the placement rule on a more conscious level, but the application is
too far off in the speech production chain.
Somewhat special are the following NEG-mains:
5. vi har läst inte så mycke (F66:I3)
(we have read not very much)
6. in min land ja kan göra inte så mycke saker (S05:I1)
(in my country I can do not very many things)
7. man kan studera inte så mycke (S05:I1)
(one can study not very much)
8. dom hade haft inte så många kompisar innan (P53:I1)
(they had had not so many friends before)
9. de har blivi inte så mycke svårare (S55:I3)
(it has become not very much more difficult)
10. men dom vill prata ibland inte (F03:I3)
(but they want to talk sometimes not)
In Examples 5-10 and 3 above the non-finite verbs are postponed by ibland inte (sometimes not) and inte så
mycket/många/adjective (not very much/ many/adjective). In main clauses containing simple verbs this type of sequences
follows the Swedish norm, e.g. han läser inte så mycket (he reads not very much), and they may also occur in verbless phrases,
e.g. nä inte så mycke (no not very much). Although we do not know exactly how frequent such sequences are in everyday
Swedish, a reasonable guess is that they are common and needed to 'grade' utterances in communicationa need that is very likely
felt by the learners, too. Consequently, such frequent and useful sequences may therefore be acquired early. Since the words are
mostly placed in the same order, they may be perceived as units and used as such in the learners' own production. Later, this
seems to restrict the application of the rule for the placement of NEG, when there is a demand for splitting the sequence. (See
also the section on subordinate clauses below.) Thus, in these cases NEG is not linked to the verb but to a sequence of other
words. However, not all learners produce errors of this kind.
The following errors may also be a result of early phrasal learning
11. men till exempel Mariehamn dom inte prata finska (F03:I1)
(but for example Mariehamn they not speak Finnish)
12. ja inte förstår (S05:I1)
(I not understand)
This type of verb must be frequent in early second-language learner communication, when the learners still have not discovered
the Swedish placement of NEG in main clauses. Subsequently, such constructions may be fossilised as phrases with pre-verbal
NEG. The data show six such incorrect

< previous page page_105 next page >


< previous page page_106 next page >
Page 106
NEG-mains with prata, and for three of the learners these clauses are the only incorrect NEG-mains in the context of V. These
learners have all been in Sweden for more than two years. Concerning inte förstår in Example 12 this subject has no other
preposed NEG than three occurrences with inte förstår during I1.
The following clauses might be seen as an extension of the phrasal processing explanation suggested here (/ marks a short
pause):
13. men flickan/hon har att märkt inte den/hunden (P52:Pil)
(but the girl/she has to noticed not that/the dog)
14. men min barn kan inte ut lek/kan gå inte ut leka (S16:I3)
(but my child can not out pla/can go not out play)
15. du måste säjer inte de (S16:I3)
(you must say not that)
16. hon vill ha inte någonting (S54:I1)
(she wants to have not anything)
In Examples 13-16 and 2 above, there are other elements after the nonfinite verbs and the negations may be seen as a negation
of constituents. Particularly interesting is Example 14 in which the speaker firstly says kan inte, but in the combination mV (kan
gå) she places NEG after the verb group. This might be seen as an example of (1) a readiness to produce kan inte with
postponed NEG; (2) a preference for keeping the later uttered verb group together; and (3) a preference for placing NEG
immediately before the intended negated element.
To summarise, most of the subjects make only few errors in the NEG-mains and the errors shown may be multi-interpreted.
There are tendencies to produce negation of constituents instead of sentence negation and for phrasal learning and phrasal
processing to take precedence over the application of syntactic rules.
Negation in subordinate clauses and 'att' (that) clauses
In this section the results of the study on subordinate clauses containing NEG (NEG-subs) are presented. These are compared to
the results from the NEG-mains and the att clauses.
The pre-verbal placement of NEG in subordinate clauses suggests that a learner who has not internalised the placement rule for
the main clauses may produce correct subordinate clauses for 'wrong' reasons, i.e. s/he uses the unmarked (and in this case also
the first language-like) placement without knowing that this is one of the markers for subordinate clauses. In order not to bias
the results NEG-subs uttered by subjects producing NEG-mains with pre-verbal NEG are excluded. This is the case for two of
the subjects.
Most NEG-subs are used by PH and SH and least by SL. Seventeen of the subjects produce no NEG-subs at all and 17 of the
others produce no NEG-subs at Time 1. Five of the high level learners notably increase their use

< previous page page_106 next page >


< previous page page_107 next page >
Page 107
of NEG-subs at Time 3, and two of them also seem to have learnt the NEG-sub placement rule. Thus, as expected, the subjects'
acquisition of the rule for NEG-subs occurs later than that for the NEG-mains.
The most frequent context in the NEG-subs for placement of NEG is V (22/95) (see Table 1). For the rest of the verbal contexts
there are very few occurrences of negation and little difference between groups. In Table 3 this is displayed for the low and high
level learners, respectively. The table shows low frequencies for some of the contexts and there is no clear picture of which
contexts may be favourable for the placement of NEG. The results for both the high and low level learners, however, indicate
that COP is a favourable context in NEG-subs too, and for the high level learners this is also the case for PERF. On the other
hand, the V-context is more favourable than the M-context. Thus, in this respect the results may be interpreted as a support for
Hyltenstam's result (1978) that V is a more favourable context for the placement of NEG in subordinate clauses than Ma
tendency just the opposite to NEG-mains.
Table 3. Different contexts in NEG-subs produced by low and high level
learners, respectively
Learner level V COP M mV PERF VV
Low 2/30 1/9 0/1 0/7 0/1 0
0.07 0.11 0 0 0 0
High 20/65 5/11 1/8 5/15 4/8 0/3
0.31 0.45 0.13 0.33 0.5 0

As in NEG-mains the combination inte så mycket (not very much) seems to be resistant to splitting and in these data there is no
correct NEG-sub containing inte så mycket but 10 error clauses, for example
17. om man skriver inte så mycke (P64:I3)
(if one write not very much)
18. som vill bli fotograferad och in- har inte så snyggt va (P53:Pi3)
(who wants to have herself photographed and no- has not very good)
Note especially Example 18 in which the speaker alters his utterance, perhaps because he conceived inte så snyggt to be a
phrase not amenable to splitting. This learner's other NEG-subs have pre-verbal NEG in the V-context and post-verbal the finite
verb in the mV- and VV-contexts. The learner (P51) who has notably more NEG-subs than the others (12/15) is also of interest
in this connection. Of three error NEG-subs, one contains the phrasal verb tycka om (like)a well known difficulty in learning
Swedishand the other two inte så (not very) in a V-context.

< previous page page_107 next page >


< previous page page_108 next page >
Page 108
This tendency to avoid splitting certain 'formulaic' expressions also holds for the att (that) clauses. Concerning these clauses,
which comprise att-, därför (att) (because) and så (att) (so that) clauses, the results give support to the hypothesis that the
placement of NEG in these clauses, due to a more variable input, might be less pre-verbal than in NEG-subs. This is especially
the case for the high level learners who are those who have best acquired the rule for NEG-subs. This is displayed in Table 4. It
can be seen from these results of the interview speech that the relative frequencies for the high level learners are lower for the
att clauses. This tendency to produce fewer pre-verbal NEGs in att clauses than in NEG-subs is obvious for some of the
learners with high proficiency in Swedish and they produce only or almost only correct NEG-mains and NEG-subs with pre-
verbal position of NEG and att clauses with post-verbal NEG.
Table 4. NEG-subs and att clauses containing NEG
Learners FL PL SL FH PH SH
NEG-subs 0/16 2/17 1/11 5/12 13/45 13/34
0 0.12 0.09 0.42 0.29 0.38
att-subs 0/10 1/26 4/13 1.25 4/34 2.27
0 0.04 0.31 0.04 0.11 0.07

Thus, there are some differences that might support the hypothesis that variable input delays the learning of a syntactic rule and
that input is as important as instruction for learning syntactic rules. Of course, anotheralthough for learner language less
plausibleexplanation might be that these learners apply the same rules for these clauses as do the native speakers, i.e. a
systematic variation between word order for main clauses and subordinate clauses depending on information structure. Another
explanation of the differences could be first language interference.
Adverbs in medial position
The study on the ADV in a medial position common to that of NEG was intended to investigate whether the variable placement
of ADV in Swedish and/or the differences in the subjects' native languages affect acquisition. With regard to ADV placement,
Finnish has rules similar to the Swedish ones for main clauses, while Polish has pre-verbal position for ADV, too, and Spanish
has variable rules but no ADV position between a finite and a non-finite verb.
Since the frequencies for ADV occurrence in contexts other than main clauses are low, only the results from the main clauses
(ADV-mains) will be presented here.

< previous page page_108 next page >


< previous page page_109 next page >
Page 109
The ADVs used by the subjects are certain time, room and attitudinal ADVs and most of them are used by the high level
learners. This is particularly true for PH whichin comparison to the other groupsuses remarkably many ADV, especially
'attitudinal disjuncts' (Quirk and Greenbaum, 1978) (74 compared to 26 for FH which is the group that comes next), whereas SL
has only one occurrence of this category. The ADV-mains are compared to the NEG-mains from the interviews in Table 5
below. In all groups the ADV-mains have lower values for relative frequency than the NEG-mains, indicating that the ADV-
rule is learned later. The smallest difference is that for FH and the largest for SH, groups that have almost equal results for NEG
(see Table 1). These results are probably due to differences in the native languages of the two groups. However, the results for
FL show similarities to those of PL and PH, which indicate that, although the Finnish and Swedish placement rules for ADV-
mains are alike, the learners do not directly 'transfer' their first language rules into the second language although their rate of
learning is higher than for the other groups. Further support for causal factors other than first language influence is the fact that
even the Finnish learners place ADV preverbally in V-context.
Concerning the error types in ADV-mains we find the following tendencies for the three native speaker groups:
The Finnish speakers make fewer errors than the other groups. The errors consist of pre-verbal placement in the V-context and
post-verbal the non-finites in vV context. These occur most markedly when a clause consists of other difficulties besides the
ADV.
The Polish speakers have ADV pre-verbal in 22 of 27 error clauses and the placement of ADV is post-verbal a non-finite verb
in five occurrences.
The Spanish speakers have a pre-verbal position for ADV in all error clauses except one, which is a vV-context.
In summary, the results indicate that these learners, despite a focus on the similarities between the placements of NEG and ADV
in teaching, do not perceive NEG and ADV as following the same rules, probably due to the variable input. For the Finnish
group, first language rules may also play a
Table 5. Main clauses containing NEG and ADV, respectively
Learners FL PL SL FH PH SH
NEG 189/201 197/202 166/187 210/214 274/275 308/311
0.94 0.98 0.89 0.98 0.99 0.99
ADV 18/22 21/28 5/8 52/54 62/82 22/35
0.82 0.75 0.63 0.96 0.76 0.63

< previous page page_109 next page >


< previous page page_110 next page >
Page 110
part. This is indicated by earlier acquisition of the ADV rules compared to the other groups. However, the similarities between
the Finnish and the Swedish main clauses containing ADV do not result in earlier acquisition of the ADV rules than the NEG
rules.
The different data types
A comparison between the learners' results in the three data types, free speech in interviews, elicited speech in picture
description and judgements of written sentences, show some interesting tendencies, though these are neither large nor
systematic.
A comparison of picture talk (Pi) with interviews (I) shows that most of the subjects make more errors in NEG-mains during Pi
than during I (because of low frequencies in Pi the NEG-subs and ADV-clauses are excluded here). This may seem a little
surprising, since one would perhaps expect that the errors would decrease following a greater focus on language form during the
describing of pictures than during the more communicative task in the interviews. Yet, a closer study shows the results to be
reasonable, as most of the errors occur when the subjects lack necessary vocabulary or are uncertain of a picture's depiction. A
possible explanation for this may be that in learner language the application of a syntactic rule collapses when processing
capacity is overloaded by searching for words.
However, most of the subjects make fewer errors during Pi at Time 3, i.e. after the Swedish course. At this time, they have quite
likely enlarged their vocabulary and become more conscious of the placement rules. Yet, the results are not that simple, since
fewer errors during Pi are also found for the subject who seems to have the least developed competence in Swedish. At Time 1
he produces NEG in full clauses only during Pi and at Time 3 his only error-free NEG-clauses are produced during Pi. He
apparently focusses more on form during the picture description task, where he also adjusts himself grammatically, than during
the interviews where he argues politically and is very eager to convey his message. Thus, linguistic behaviour in different
situational contexts is very likely a question of focus due to personality.
The acceptability test includes written sentences containing NEG and ADV that the subjects are asked to judge for correctness.
The sentences are two NEG-mains, six NEG-subs (two relative, two prepositioned and two postpositioned temporal clauses) and
three ADV-subs. Four of the sentences are correct and the rest incorrect with regard to the placement of NEG or ADV. The
results show that the subjects' ability to estimate the correctness of sentences mostly agree with their production in speech.
Concerning the ADV-subs the results for the Finnish-speaking group show a preference for post-verbal placement of ADV and
their number of error judgements is substantially larger.
In addition, the test items indicate that it is apparently much easier to consent to a correct sentence than to object to (and
correct) an incorrect

< previous page page_110 next page >


< previous page page_111 next page >
Page 111
one. Thus, the construction of a test seems to be of great importance for the results gained.

Concluding Remarks
On the basis of these data the answer to the question in the title of the paper will be 'Yes, there is some order'. For these tutored
subjects the rules for the placement of NEG are learned earlier than those for ADV, the auxiliary contexts in main clauses
earlier than the main verb contexts and there is a weak tendency for the opposite order in subordinate clauses. In these respects
all learner groups and mostly all learners behave in the same way. Concerning the placement rules for ADV there is an obvious
tendency for the Finnish high level subjects to learn the rule for main clauses earlier than the other groups, which is probably
due to effects from the native language. Their preference for post-verbal placement is particularly obvious in the subordinate
clauses in the acceptability test.
Other kinds of order indicated in the results are the tendencies (1) for semantics to take precedence over syntax and (2) for
prefabricated patterns to take precedence over syntactic rules in speech production.
By the first tendency (1) it is understood that learners neglect a syntactic rule under acquisition when problems arise in
expressing intended meaning. This is most obvious during the picture description task, when sometimes the learners lack the
necessary vocabulary, i.e. a situation with no possibilities of avoidance. In these cases the learners return to an earlier 'stage' in
the development of the placement rule (mostly pre-verbal placement in main clauses).
The second tendency, (2), accounts for many learners' preference to process frequent sequences as prefabricated patterns,
thereby rejecting syntactic rules that are applied in other contexts. This preference is found both in main and subordinate clauses.
In fact, the learners' early acquisition of frequent sequences seems very plausible, because sequences that are frequent in
everyday communication are, of course, expressions that are needed in human communication and hence important in the
learners' communication, too, and therefore they are acquired early (cf. Hatch, 1978). This is probably the reason for the early
acquisition of sequences with the auxiliaries as well. Their frequent use, their simple forms (cf. Zipfs law) and their importance
in human communication cause them to be acquired early. Besides, memorising useful phrases ought to be an economic way of
learning, comparable with, for example, the way many of us memorise a long telephone-number.
Early phrasal learning may also be the explanation of the later difficulties, especially in the contexts of auxiliaries, when the
placement rules for the subordinate clauses are being acquired, and perhaps this is the reason for the placement of NEG after the
non-finite verb in verb groups, too. The perfect and past-perfect tense and the modal plus the main verb may be

< previous page page_111 next page >


< previous page page_112 next page >
Page 112
drilled in teaching in one way or another, and hence perceived as units and therefore kept together in speech production when
the learner's main focus is on anything else but syntax. Notably, it is not always the grammatical structural units that are kept
together, e.g. the main verb and its object, but what the speaker has perceived and stored as units, e.g. inte så mycket (not very
much). Further evidence for this type of phrasal learning is *den är så mycket bra (it is so much good) which according to
teachers is a common error in Swedish learner language.
Of course, this tendency for phrasal learning and phrasal processing is not valid for all subjects and very likely there are
different ways of learning a language, of which phrasal learning is one. Further examination of the data may show if these
tendencies for many of the learners also hold for other linguistic features. Furthermore, a crucial question following the
hypothesis suggested above is how to explain the gradual transition from obvious use of phrases to use of creative syntactic
rules.
The errors may be explained by competing rules in the learners' interlanguage and seen as examples of how a rule is slowly
extended to all linguistic and situational contexts. Yet, a question here is why some learners develop their rule and some do not
during the course. Background data for these learners give no explanation. There are no systematic differences between the
'developers' and the 'non-developers' with regard to variables such as age, length of stay in Sweden, earlier education,
possibilities of speaking Swedish outside the course or the learners' intentions to stay in Sweden.
Acknowledgements
This study is part of the project Språkutveckling och undervisningsmodeller (SUM) which has been financed by the Swedish
National Board of Education. I would like to thank Christopher Stroud for valuable comments on my English and Pirkko
Forsman-Svensson and Rita Kozlowska-Ras for valuable information about Finnish and Polish, respectively.

References
Andersson, L. G. (1975) Form and function of subordinate clauses. Gothenburg Monographs in Linguistics. University of
Göteborg: Dept. of Linguistics.
Bolander, M. (1987) Man kan studera inte så mycke. Om placering av negation och satsadverb i vuxna invandrares svenska.
SUM-rapport 5. Stockholms universitet: Institutionen för lingvistik.
Bourgonje, G. C. J. (1984) (Non-) language-specific deviations in EFL adverbial usage. Riksuniversiteit Utrecht: Institut voor
Engelse Taal en Litteraturkunde.
Cazden, C. B., Cancino, H., Rosansky, E. J. and Schumann, J. H. (1975) Second language acquisition sequences in children,
adolescents and adults. Final report United States Department of Health, Education and Welfare.
Clahsen, H. (1984) The acquisition of German word order: a test case for cognitive approaches to L2 development. In R. W.
Andersen (ed.), Second Languages: A Cross-Linguistic Perspective. Rowley/Mass.: Newbury House.
Clahsen, H., Meisel, J. M. and Pienemann, M. (1983) Deutsch als Zweitsprache. Der Spracherwerb ausländischer Arbeiter.
Tübingen: Gunter Narr.

< previous page page_112 next page >


< previous page page_113 next page >
Page 113
Dahl, Ö. (1979a) Typology of sentence negation. Linguistics, 17, No. 12, 79-106.
Felix, S. W. (1978) Linguistische Untersuchungen zum natürlichen Zweitsprachenerwerh. München: Wilhelm Fink Verlag.
Givón, T. (1984) Syntax. A Functional-Typological Introduction. Philadelphia Amsterdam: John Benjamins.
Hakuta, K. (1974) Prefabricated patterns and the emergence of structure in second language acquisition. Language learning, 24,
287-97.
Hammarberg, B. (1979) On intralingual, interlingual and developmental solutions in interlanguage. In K. Hyltenstam and M.
Linnarud (eds), Interlanguage. Workshop at the Fifth Scandinavian Conference of Linguistics, Frostavallen, 27-29 April 1979.
Stockholm: Almqvist & Wiksell.
Hatch, E. (1973) Second language learninguniversals? Working Papers on Bilingualism, 3, 1-18.
(1978) Discourse analysis. In E. Hatch (ed.) Second Language Acquisition. A Book of Readings. Rowley/Mass.: Newbury
House.
Horn, L. R. (1978) Some aspects of negation. In J. H. Greenberg (ed.) Universals of Human Language. Stanford: Stanford
University Press.
Hyltenstam, K. (1978) Variation in interlanguage syntax. Working Papers No. 18. Lund: Department of General Linguistics,
Lund University.
Klima, E. S. and Bellugi, U. (196) Syntactic regularities in the speech of children. In J. Lyons and R. J. Wales (eds),
Psycholinguistics Papers. Edinburgh: Edinburgh University Press.
Lindberg, E. (1973) Studentsvenska 1864-1965. Stockholm: Skriptor AB.
Payne, J. R. (1985) Negation. In T. Shopen (ed.) Language Typology and Syntactic Description: Vol. 1. Clause Structure.
Cambridge: Cambridge University Press.
Quirk, R., and Greenbaum, S., (1978) A University Grammar of English. A Grammar of Contemporary. London: Longman.
Ravem, R. (1978) Two Norwegian children's acquisition of English syntax. In E. Hatch (ed.), Second Language Acquisition. A
Book of Readings. Rowley Mass.: Newbury House.
Sennekamp, M. (1979) Die Verwendungsmöglichkeiten von Negations-zeichen in Dialogen. München: Max Hueber Verlag.
Wode, H. (1981) Learning a Second Language. Vol. 1: An Integrated View of Language Acquisition. Tübingen: Gunter Narr
Verlag.

< previous page page_113 next page >


< previous page page_ii next page >
< previous page page_115 next page >
Page 115

Some Psychological Aspects of Early Second-Language Acquisition


Mirjana Vilke
Faculty of Philosophy, English Department, University of Zagreb,
Salajeva 3, 41000 Zagreb, Yugoslavia.
Abstract. In this paper some results of a study of children's acquisition of foreign languages (in this particular case
English) at an early school age will be discussed. The study on which the paper is based has been conducted in formal
classroom situations with children aged eight being exposed to English for a couple of periods per week and being
surrounded by the language and culture of their native country. Consequently, this is a case of foreign language learning
rather than second language acquisition.
The focus of the analysis is the motivation of the learners and the sources of difficulties during the learning process.
The advantages of an early start with a foreign language are discussed: the familiarisation with the linguistic properties of
the foreign idiom at an age when it is readily accepted; the beneficial effect upon the cognitive growth of the individual;
and the prevention of the development of ethnocentric tendencies in later life.
Since 1984 when the first generation of 'early starters' reached the final year of their primary school, their attitude to
English has been observed and compared to the control groups who started English at a later age. In the paper some
findings of the observations are reported.

Introduction
The ultimate aim of introducing a foreign language into the school curriculum while the learner is very young is the production
of competent bilingual speakers throughout the population.
Our immediate concern is with the initiation of the process: when and how to start a foreign language to ensure a sound basis
for the accomplishment of the goala bilingual speaker. For, appealing though the idea of a community of bilingual children
might be, we are, for reasons of economy,

< previous page page_115 next page >


< previous page page_116 next page >
Page 116
interested in it only marginallyas an eventual by-productwhile our main target is the bilingual adult capable of using a foreign
language in his life and vocation.
The reasons that speak in favour of an early start seem to stem from three different sources:
the findings of neurophysiology, and developmental psycholinguistics;
empirical and experimental evidence derived from the results of acquiring a second language by groups and individuals through
a variety of immersion programmes organised in the countries where the target language is indigenous;
empirical and experimental evidence derived from the results obtained by formal classroom tuition in different sociocultural
settings in the learners' mother-tongue environments.
The point to be made with regard to neurophysiological evidence is that much of it is derived from the study of the abnormal, as
Roberts has already cautioned (1973: 105). Developmental psycholinguistics, on the other hand, observes the behaviour of
children acquiring their first language (L1), so neither of these investigations is directly connected to the acquisition of a second
language (L2). Nevertheless, a number of hypotheses (highly relevant for the study of acquisition of an L2 at an early age) about
the linguistic and intellectual development of the child can be drawn from the evidence collected both by neurophysiology and
developmental psycholinguistics.
Experimental and empirical evidence from different programmes conducted in the L2 country, used in designing the
programmes for the formal classroom teaching in the learner's LI country, can be vastly misleading, as a number of
environmental and operational factors are as a rule entirely differenttime of exposure to L2 and the sociocultural setting in
which learning takes place being the most obvious ones. But once the differences are identified, evidence from such
programmes can be utilised in classroom situations, especially when it offers insights into the process of the acquisition of an
L2, which seems to havefrom the little we know about it at presentcertain universal features.
As regards experimental and empirical evidence from the results of teaching an L2 in classrooms in the learners' own
countrythis should not be generalised before the conditions in different programmes have been identified and verified, the
variables being so different. Differing motivation and attitudes of learners towards a particular L2 can account, for example, for
the differences between the results obtained.
The only logical conclusion that can be inferred from what has been said is that, valuable as it is, evidence stemming from any
of the sources must be treated with caution and regarded as raw material which should be further processed and investigated
rather than used as a premise on which to build

< previous page page_116 next page >


< previous page page_117 next page >
Page 117
theories about L2 acquisition at an early age. And if it ever comes to theories it will probably be possible to prove them only in
the specific social, linguistic and cultural setting in which the research has been conducted.

The Zagreb Project


In 1973 a project was started in Zagreb with the aim of providing evidence of the characteristics of the process of learning
English at an early age in formal school environments.
In the first stage, completed in the school year 1973-74, research was conducted to find out to what degree (if any) children
learn English with more ease before puberty than learners who have passed Lenneberg's 'critical period' (1967). Sixty beginners
aged nine and 60 beginners aged 17-19 were supplied with the same language material, delivered by the same method and
approximately the same techniques of work, for the same period of time. Care was taken that both materials and techniques used
should suit learners of both age groups pedagogically, an almost impossible task.
The results proved that there were differences between the pre- and post-puberty groups. The post-puberty group had certain
advantages, such as more insight into the functioning of language, the experience of studying their mother tongue and some
other language, intellectual maturity, and so they were faster learners of structures and vocabulary The pre-puberty group were
far superior in mastering the phonetic system. On the level of pronunciation the most striking differences were noticed between
the two groups: the older group as a rule used mother tongue approximations of English phonemes. The deviations from the
norm were such that they sometimes blurred the meaning of utterances. The younger group reached a considerably higher
standard of pronunciation, using authentic English phonemes and intonation patterns in most cases.
The findings of the investigation were consistent with Lenneberg's statement about 'language learning blocks' which become
frequent after puberty (Lenneberg, 1967).
The results of the investigation led us to conclude that learners should start foreign languages well before puberty. This would
provide them with sufficient time to acquire a good command of the phonetic system with a limited corpus of structures and
vocabulary, and provide them with a feeling of security and self-confidence as regards the foreign language. Once they pass the
age of the 'maturation of the brain' they will be able to proceed to more subtle and abstract uses of the foreign language (Vilke
1975, 1976a).
The second stage of the Project started in 1975. The aim of the investigation was to find out at which age between six and nine
it would be best to start a foreign language at school, and what factors play a role in the learning process at this age.
Seventy children aged six to nine were observed during a year-long English

< previous page page_117 next page >


< previous page page_118 next page >
Page 118
course at a language school in Zagreb. The results provided the tentative answer that eight to nine would be the optimum age at
which to start English for most children, and that six to seven would suit those of above-average intelligence. But these results
are valid only in this particular sociocultural environment in Zagreb and should not be generalised without further research. The
children in question came from middle-class families who had sent their children on an English course for a variety of reasons,
one of them being the realisation of the need to be able to communicate in a foreign language (especially English), and another
'to keep up with the Jones's' whose children also study English.
As this stage of the Project was conducted in a specialised foreign-language school where the courses are not part of the
compulsory educational system provided by the government for the entire population, it was evident that the motivation to study
English (which by no means could have been expected in this age group, anyway) came from parents. The parents' attitude was
positive, and this was helpful in the initial stages. During the course of study, however, the children developed their own
attitudes towards English which very largely depended on their personal attitude towards the teacher. It was observed that the
child's feeling at ease, and a positive emotional link with the teacher, accounted for the greater part of the success of individual
children.
Children progressed through the language corpus making their own discoveries of its system (for example 'It must be cars for
three'). They were guided through it by the teaching materials designed to provide appropriate context for the five functors
found by Dulay and Burt (1973) to be internalised first by groups of Spanish and Japanese children acquiring English in the
U.S.A.
Unfortunately, the 'natural order hypothesis' worked only where there was no negative transfer from L1. Both interference and
developmental errors were observed, which was not in keeping with one of Dulay and Burt's statistics in which they found only
3% interference errors (Vilke, 1976b). So the empirical evidence has supported the view that acquiring a second language and
learning a foreign language are different processes, the confusion of which can be a dangerous outcome.
The third stage of the Project started in 1977. English was introduced for children in the second grade (eight years old) in five
primary schools in Zagreb, on an experimental basis. According to accepted Yugoslav standards a foreign language is as a rule
introduced in the fourth grade, when learners are ten years old.
The schools chosen for the experiment were situated in suburban areas in which there was little tradition of communication in
any foreign language. On the other hand the situation in these schools would resemble fairly accurately the situation in any
suburban area or village in the country, and therefore should be valuable as a pilot experiment prior to the introduction of foreign
languages into the second class on a large scale.

< previous page page_118 next page >


< previous page page_119 next page >
Page 119
Motivation
Before we started any teaching we tried to obtain certain clues about the attitude of our students to things English. Every child
was interviewed by a member of the Project, but not his future teacher.
Children were asked 13 questionsnine of them connected with the child's idea of English and the English, and four in connection
with their parents' and friends' attitudes to English. Gardner & Lambert have in the course of their studies proved that 'parents
who are instrumentally oriented appear to pass their orientation on to their children' (Gardner & Lambert, 1972: 128). We could
not have expected to have parents motivated with respect to their children studying English, but we feared a negative attitude on
their part which could have caused very serious damage to the entire undertaking.
The analysis of the interviews showed the following attitudinal characteristics of our potential learners:
1. They were looking forward to the English classes.
2. They expected routine school work.
3. A minority of students realised that learning English could be useful. As a rule they displayed instrumental orientation
towards it.
4. They did not show much desire to visit English-speaking countries.
5. The subjects gave a relevant opinion of English spoken in our country.
6. In the children's families there was no tradition of cultural influences coming from English-speaking countries. Parents were
fairly indifferent to their children studying English; they did not encourage it, but neither did they oppose it.

All this led us to conclude that the task ahead was a very responsible one. The raw material we had to mould was in the hands
of the teachers. If they managed to motivate children to study English at that early age it might have a life-long beneficial effect
not only on their ability to use English but also upon their attitude to foreign languages and other nations in general.
One school year should suffice to find out if the learners have become used to the new subject and developed their own attitudes
towards it.
Each group of 12-15 students had two periods of English per week on a regular basis. Another interview on motivation was
conducted at the end of May 1978. The objective was to find out whether attitudes towards English classes had changed, and, if
so, in what direction. This time they were given nine questions, from which a picture of their likes and dislikes could be formed.
It was considered important to hear their interpretation of what they like, and what they find difficult in English lessons. The
parents' attitude was tested again to find out whether it had changed in the course of the year's study.

< previous page page_119 next page >


< previous page page_120 next page >
Page 120
Analysis of the second interview answers indicated that a year-long teaching of English changed the children's attitude of
moderate curiosity to a desire to proceed with learning English.
1. Approximately 70% of the children liked English classes because the content was adjusted to their interests and because
they felt free and encouraged by the teacher. Thirty percent of the children enjoyed them because they felt they were getting
somewhere in their attempt to learn how to communicate in English.
2. The attitude of the parents changed from one of general disinterest, to encouragement of children to persevere.
One general conclusion that had been drawn from both interviews was that even in environments where there was no positive
orientation towards a foreign language, children could be motivated to study it if they were approached in the right way, and
their motivation could influence their parents' attitudes towards it. This, in turn, could have a beneficial effect on the
international orientation of the whole community. In our particular case the process was somewhat different from the one
observed by Gardner & Lambert in which 'parents with positive attitudes towards the other language community more actively
encouraged their children to learn that language' (Gardner & Lambert, 1972: 6), as here the motivation of the children caused a
change of attitude in the parents.
Van Parreren (1976) argues that an early start in teaching foreign languages could create motivational problems; learners would
have to spend such a long time on learning that they would lose enthusiasm for the language. A later start and a more
concentrated effort could avoid problems of long-range motivation. However, this is not a convincing argument. The basic idea
of introducing English into school curricula on an experimental basis at an early age was to motivate the learners to make use of
the language. It seemed only natural that the more familiar they become with it and the better they can manipulate it, the more
willing they are likely to be to use it.
In 1984 the first generation of our 'early starters' reached the final year of their primary school. So far we have been able to
observe four generations of them. Their motivation and attitude to English were examined and compared to a control group who
had started English in the fourth grade.
The results of the analysis have not been completely processed yet, as the last interviews were conducted only recently, but
some rough estimates as to the differences in motivation between the experimental and control groups can be offered. In
following years we plan to continue the investigation to see whether the learners differ generationwise. All the results will be
submitted to statistical processing, of course, and published in a detailed report.
Here are some of the preliminary findings of the latest investigation into the learners' attitudes and motivation, which will show
the basic differences between experimental and control groups. It should be stressed that the

< previous page page_120 next page >


< previous page page_121 next page >
Page 121
differences in motivation between the 'early starters' and the control group are not as strongly pronounced as we expected them
to be.
The learners are, on the whole, motivated to study English and they have positive attitudes towards it. The number of those who
do not like it is slightly larger in the control group. There are more students in the control group to whom English seems to be a
problem subject. In the control group the learners are more afraid of bad marks in English. The control groups feel a greater
need for more weekly periods of English, and in some of them English is considered significantly more difficult. The
experimental groups on the whole consider English slightly more interesting, whereas in the control groups they stress more the
useful side of studying it.
Apparently the early start influences positively most of the students' attitudes to English. However, in the course of the years so
many parasitical factors enter the process of learning that it is difficult if not impossible to differentiate among them. There have
been some interesting by-products of the investigation. The location of the school from which the learners come seems to be
responsible for the differences in motivation even more than the age at which they start: learners coming from a certain areaboth
experimental and control groupsare significantly more highly motivated to study English than their peers coming from another
suburban area. This was not due only to the personality of the teacher they shared although this has proved to be an extremely
important factor. But even where they have different teachers in the same area, they showed similar trends in their attitudes to
English. It may well be that the influence of the local school by and large, or the local cinema and other mass media, are
responsible for that. In future analyses we shall try to examine more profoundly a set of social factors that are likely to influence
the learners' attitudes to the English language and things English.
Sources of difficulties
Performance difficulties in the process of learning a foreign language can be expected on the level of pronunciation, command
of structures and use of vocabulary, if we take into consideration a narrow linguistic (and not a wider, sociolinguistic) aspect of
performance. This is the aspect that will be discussed here, as the subjects of our investigations are for the time being limited to
classroom performance and the linguistic content of the course. (This content is communicatively based but most learners will
not have a chance to test its effectiveness in real life in the immediate future.)
In our Motivation Interview, only 9% of the students claimed that they had difficulties with pronunciation. A guess could be
ventured that the percentage of adult beginners who have difficulties with pronunciation would be much higher.
Observations of the children's performance have proved once again that they can master the phonetic system of the foreign
language provided they

< previous page page_121 next page >


< previous page page_122 next page >
Page 122
have good models to imitate. Accordingly, complete accuracy in pronunciation, rhythm and intonation patterns should be
insisted upon with no fear of overloading the learners, who should make the most of the advantages their age offers. It should be
remembered that one of the reasons for introducing a foreign language at this early age is to familiarise the learner with a
pronunciation system different from his own, at an age when he does not feel threatened by it.
Contrary to the case observed in the mastering of the elements of English pronunciation, it has been observed that some words
are learned with ease, whereas other words present difficulties. In the second interview on motivation, 32% of the children
complained about difficult words.
In an experiment specially devised to test this phenomenon we opposed two pairs of words, the first pair being bottle-kettle, and
the second television-fireplace. The learners from three of our five experimental schools were included in the experiment80
children altogether. The vocabulary items were practised over the same period of time, and more or less identical techniques and
aids were used. Neither children nor teachers were informed of the content of the experiment beforehand, so they could not have
had any additional practice. The subjects were shown a bottle and a kettle and asked what the English words were for the
objects. Fifty-six examinees were questioned on the kettle-bottle pair; one class did not have sufficient practice beforehand, so
their answers were not accepted. Thirty children remembered the word bottle and only seven remembered kettle. It was repeated
with the words television and fireplace. Out of 80 children, 74 could use the word television, and only 22 fireplace. Kettle and
bottle were selected because on the phonological level they are similar; in both, the same consonant cluster /tl/occurs, and
neither consists of phonemes that do not appear in our own phonological system. Fireplace and television were selected as both
words are compounds; neither presents any special phonological difficulty and their length is approximately the same. However,
teachers reported difficulties in practising the words fireplace and kettle. What is the source of this difficulty? We hypothesised
that it was due to the fact that neither fireplace nor kettle presents concepts familiar to the children. (In this country, homes are
rarely heated by means of fireplaces, and a kettle is not a common object in most households as we are a nation of heavy coffee-
drinkers.) The teachers found the same sort of difficulty with words like mantelpiece, chest of drawers and many others that
would be perfectly simple for a child coming from a British cultural background.
This finding was, we thought, helpful in two ways: first, it has contributed a little to our understanding of the way children learn
a foreign language; they seem to transfer the concepts they have acquired in their L1 into their L2, L2 being a foreign and not a
second language. In this particular case they re-name the spontaneous concepts with foreign names if they are identical, and
have to develop non-spontaneous concepts (which are, according to Piaget, influenced by adults) if the concepts are non-existent
or

< previous page page_122 next page >


< previous page page_123 next page >
Page 123
different in their L1 culture, this being a more complex process. One could call it a negative transfer from L1, on the conceptual
level. This finding does not necessarily contradict Hernandes-Chavez' opinion that,
The bilingual learner acquires two distinct (though obviously very closely related) semantic systems, i.e. he proceeds
under basic assumptions that the new language may express quite different semantic functions. (Hernandez-Chavez, 1977:
149).
The key to the difference should obviously be looked for in the difference between the second and the foreign language; that is,
the difference in the environment and the intensity of the exposure to an L2. Hernandez-Chavez' examinee was a three-year-old
Mexican child exposed to English among native English peers day after day in a day-care centre in California. It is only logical
that the child should in such circumstances build up a semantic system parallel to the one in his mother tongue. No analogies
with such cases should be made with our learners exposed to English for two periods per week, among L1 peers and teachers.
The implications of our finding could be applied by those who design teaching materials for young school children; objects
denoting concepts not familiar to the children should (if at all possible) be laid aside for use at a later age and a later stage when
their introduction may be welcome as another interesting detail revealing new data about the L2 peoples' ways of life.
In the same experiment, several items were designed to probe children's ability to comprehend and produce structural categories.
They were tested on the production of plural forms, the comprehension of spatial relations expressed by the preposition on, and
on pronominalisation. A few characteristics of children's use of structural items emerged as by-products of the investigation.
Test item on plural
In front of them there was one apple, and a little further away three apples placed next to each other. They were asked to name
one apple first (to recall the word) and three apples after that. Three variants of the answer were accepted as correct. Apples,
three apples, these are three apples. Only 27 out of 80 children offered a correct answer, but 95% (76 out of 80) children used
one signal of plural (three apple, these are apple) in their answer. It seems that children in their process of learning a foreign
language understand the concept of plurality, they feel it must be marked, and they mark it, but leave out whatever is (to their
mind) redundant.
A practical hint for teachers would be that they should not insist on all the plural markers when teaching this age group, and that
they should be content with some signal for plurality if the child wishes to convey his own thoughts and ideas. Intensive practice
of the correct forms will come at a later stage, when the child's mind works more systematically.

< previous page page_123 next page >


< previous page page_124 next page >
Page 124
Test item on the preposition on
They had to perform two commands to show their comprehension of on. After the bottle and the kettle had been identified, the
children were told: put the bottle on the floor, put the kettle on the chair. Both commands were performed correctly by 94% of
the population, which obviously shows that they understand the relations expressed by on.
Test item on pronominalisation
Children are reluctant to use the pronouns he and she and they much prefer using nouns. The task of the test was to either
confirm or reject the observed characteristic of children's speech as a regularly occurring pattern. In the test they were presented
with the picture of a boy with a red ball and the picture of a girl with a blue flower. The examiner asked 'Who has got the red
ballhe or she?' and, 'Who has got the blue flowerhe or she?' Eighteen children used he or she in their answers, 32 children
answered boy or girl, and 6 children confused the pronouns. Only 56 answers were accepted, as in one class the teacher who
was present was trying to help and so influenced the children's answers. These answers could not be regarded as spontaneous,
and they were not accepted.
The conclusion of the test, prompted by the observations during the lessons, is that most children understand what he and she
stand for, but prefer using nouns, which probably look less abstract. This is in keeping with Chomksy's investigation in which
she found the process of pronominalisation in English as an L1 still in the state of development in 6-7 year-old children
(Chomsky, 1969). As children learning a foreign language cannot be expected to process successfully those features of language
not fully mastered by their peers in their mother tongue, it would be advisable to postpone pronouns until a little later. In this
way the frustrations of both teachers and learners would be avoided.
In the test several features of children's speech were observed.
1. The permanent misuse of the articles, which they used at random; sometimes as part of the noun; much more often not
using them at all, or using them completely inadequatelye.g. indefinite article with the pluralthree an apple. This is probably due
to heavy interference from the mother tongue in which the articles do not exist. The children showed a complete inability to
establish a frame of reference for the articles, which Duskova * considers to be the gravest form of interference (Duskova,
1969). This characteristic of children's speech has been discussed in some other articles (Vilke, 1976b).
2. The existence of 'prefabricated patterns', which Hakuta (1974) found in acquiring English as a second language, has been
found in our case, too. He defines them as one of the possible strategies employed by learners when they wish to express
thoughts in the target language

< previous page page_124 next page >


< previous page page_125 next page >
Page 125
but do not yet know the forms. In our corpus we found Mary sit down (after the command Sit down) it's (I can see it's a cat)
I've got (This is I've got flower). They seem to be a sign that the student tries hard to express his ideas in the foreign
language.
Observations and the results of tests administered in the course of the Project seem to indicate that there are several
characteristics of children's performance in English as a foreign language that constantly recur in the course of the learning
process:
1. They can master the phonological system of English with the greatest ease.
2. Vocabulary items for which they have not developed concepts in their own culture present difficulty.
3. They can understand basic relationships in a sentence, especially spatial relationships expressed by prepositions, and the
concept of plurality, etc.
4. Difficulties in learning structural elements stem from two main sources:
a interference of the mother tongue (this can be seen in the use of articles);
b immaturity, which makes certain concepts in both the primary and secondary language hard to grasp.
5. Interference from the mother tongue manifests itself at both the linguistic and the conceptual level.

Conclusion
Our Project seems to have shown that children in our particular sociocultural environment have been able to start successfully
the long process of becoming bilingual.
None of them has yet become bilingualone could hardly expect this to happen in such a limited number of school periods of
learning English. According to an estimate (Mueller, 1967), six months, with between 6 and 14 hours a day, would be needed to
achieve proficiency in a foreign language, but only when selected students are in question. (This amounts to a total of 1,000 to
1,500 contact hours.)
Motivation to learn English, which was zero before the start, developed significantly. Ninety-nine percent of the children at the
end of the first year of learning expressed a strong desire to continue.
We hope that this has had a strong effect on moulding their life-long attitude towards English as a foreign language and foreign
languages in general, preventing the development of ethnocentric tendencies later in life. (Ethnocentrism is defined by Gardner
& Lambert (1972) as stereotyped negative feelings toward foreign countries and peoples.)

< previous page page_125 next page >


< previous page page_126 next page >
Page 126
Our young students have become acquainted with the concept of English in a way that corresponds to their ideas of 'interesting'
and 'amusing'; most associations connected with it have been pleasant; no fear of punishment (bad marks, ridicule, etc.)so often
a permanent companion of school activitieshas been present, and even feed-back of their orientation has been observed in the
changed attitudes of parents. In this respect, the Project can be said to have been a success.
So far no 'balance effect' has been observed in our learners. The balance effect is a hypothesis that the more time one spends on
the second language, the less well one learns the first language, with consequent detrimental effects on the native language, on
education and on the intellectual development of the child (Jakobovitz, 1971: 52).
The learners' intelligence was tested by verbal and non-verbal intelligence tests in the experimental as well as the control groups.
So far it has been observed that success in learning English is positively correlated with the learners' intelligence. At the end of
the first year of teaching the teachers were asked to evaluate impressionistically the achievements of their students by marks
from 1 to 5. Correlation between intelligence and success in learning was measured by the rank correlation method, and it varied
for different classes between 0.5& and 0.67, which proved to be significant.
This relatively high correlation is probably due to the fact that the learning was designed as a cognitive rather than a habit-
forming process. In Jakobovitz's well-known table showing the contribution of factors decisive for success in learning,
intelligence accounts for 20% of the variance, which is less than in our case where the children whose intelligence was lower
than average achieved very poor results (Jakobovitz, 1971: 98). But as has already been mentioned, this is probably due to the
fact that Jakobovitz had in mind a learning process based on habit-forming, which requires a smaller input of intelligence on the
part of the learners.
We have tried to show that foreign language learning is a process with characteristics of its own, and that it should be
investigated as such. It is not a mere duplication of the process of acquiring a first languae, nor is it identical to the acquisition
of a second language, although it resembles both processes in certain aspects.
There is probably more resemblance between the acquisition of a first language and the acquisition of a second language by a
child than there is between the acquisition of a second language and the learning of a foreign language by a child. The difference
is explained by environmental factors and the amount of time spent on the language.
The process of learning a foreign language as a child is not identical to the process of learning a foreign language as an adult,
either. The difference is between a developing and fully developed personality, with all the implications the differing concepts
bring about.
If the early school years are to be utilised for learning foreign languages (and there seem to be many reasons that support the
idea) these differences

< previous page page_126 next page >


< previous page page_127 next page >
Page 127
should be taken very seriously into consideration. But at present we still know very little about these processes, and the main
work of finding out more is ahead of us.

Bibliography
Boyd, G. (1976) Linguistics in the Elementary School. Peacock.
Brown, R. (1973) A First Language. Harvard University Press.
Burstall, C. (1974) Primary French in the Balance. A Report from the British National Foundation for Educational Research.
Chomsky, C. (1969) The Acquisition of Syntax in Children from 5 to 10. Chicago: MIT Press.
Cook, V. (1977) Cognitive processes in second language learning. IRAL XV/1, 1-21.
Corcoran, G. (1976) Language Experience for Nursery and Kindergarten Years. Peacock.
Dulay, H. and Burt, M. (1972) Goofing: an indicator of children's second language learning strategies'. Language Learning, 22,
235-52.
(1973) Should we teach children syntax? Language Learning, 23, 245-58.
(1974a) Errors and strategies in child second language acquisition. TESOL Quarterly, 8, 2.
(1974b) A new perspective on the creative construction process in child second language acquisition. Language Learning, 24,
253-79.
Duskova *, L. (1969) On sources of errors in second language learning. International Review of Applied Linguistics, 7, 11-36.
Gardner, R. & Lambert, W. (1972) Attitudes and Motivation. Rowley, Mass.: Newbury House.
Garvie, E. (1976) Breakthrough to Fluency. Oxford: Basil Blackwell.
Gillis, M. (1975) The acquisition of the English verbal system by two Japanese children in a natural setting. Unpublished M.A.
Thesis, McGill University.
Hakuta, K. (1974) Prefabricated patterns and the emergence of structure in second language acquisition. Language Learning, 24,
287-99.
(1976) A case study. Language Learning, 26, 321-53.
Hernandez-Chavez, F. (1977) The development of semantic relations in child second language acquisition. In M. Burt, H. Dulay
and M. Finocchiaro (eds), Viewpoints. Regents Publishing Co, 127-52.
Jakobovitz, L. (1971) Foreign Language Learning. Rowley, Mass.: Newbury House.
Krashen, S. (1973) Lateralization, language learning and the critical period. Language Learning, 23, 63-74.
Lamendella, J. (1977) General principles of neurofunctional organization and their manifestation in primary and nonprimary
language acquisition. Language Learning, 27, 155-97.
Larsen, D. and Smalley, W. (1972) Becoming Bilingual, New Canaan.
Lee, William (1977) For and against an early start. Foreign Language Annals. New York.
Lenneberg, E. (1967) Biological Foundations of Language. Chichester: J. Wiley.
Maratsos, M. (1976) The Use of Definite and Indefinite Reference in Young Children. Cambridge: Cambridge University Press.
Mueller, K. (1967) The army language school and its implications. In Levenson and Kendrick (eds), 479-85.
Penfield, W. and Roberts, L. (1959) Speech and Brain Mechanisms. Princeton: Princeton University Press.
Piaget, J. (1952) The Origins of Intelligence in Children. New York: International Universities Press.
(1954) The Construction of Reality in the Child. New York: Basic Books.
(1968) Six Psychological Studies. London: University of London Press.
(1973) Memory and Intelligence. London: Routledge and Kegan Paul.
Roberts, J. T. (1973) The LAD hypothesis and L2 acquisition. Audiovisual Language Journal, 2.
Spicer, A. (1978) The early teaching of modern languages. Paper at the AILA Congress, Montreal, August 1978.

< previous page page_127 next page >


< previous page page_128 next page >
Page 128
Stern, H. H. (1963) Foreign Languages in Primary Education. Hamburg: UNESCO Institute for Education.
(1969) Languages and the Young School Child. Oxford: Oxford University Press.
Van Parreren, C. F. (1976) The psychological aspects of the early teaching of modern languages. IRAL, XIV12, 135-43.
Vilke, M. (1975) Utjecaj lingvistickih * teorija na razvoj metoda u nastavi engleskog jezika. Unpublished doctoral thesis,
Zagreb.
(1976a) The age factor in the acquisition of foreign languages. Rassegna Italiana di Linguistica Applicata, No. 2/3.
(1976b) Implications of the age factor on the process of acquisition of an L2. SRAZ 41-42, 87-104.
(1979) English as a foreign language at the age of eight. SRAZ XXIV (1-2) 297-336. Vygotsky, L. (1962) Thought and
Language, Chicago: MIT Press.

< previous page page_128 next page >


< previous page page_129 next page >
Page 129

Immigrant Children's SwedishA New Variety?


Ulla-Britt Kotsinas
Department of Scandinavian Languages, University of Stockholm,
10691 Stockholm, Sweden
Abstract. As a consequence of the great immigration in Sweden during recent decades, about one tenth of the children in
Swedish schools have an immigrant background of different kinds. Many of them use a Swedish that differs more or less
from native Swedish and that has been characterised as 'poor' and as the result of incomplete learning. In this paper two
hypotheses are discussed: (1) Swedish as used by immigrant children may show certain features, related to a creolisation
process; and (2) the Swedish language may in future show signs of influence from the varieties used by persons with
immigrant background.

Introduction
In the following translated composition, written in Swedish by a Turkish girl living in Sweden, a situation is described that is not
unique in Sweden today:
My classmates
Our class is an unusual class. There are only three Swedes in the class. We speak Turkish, Serbocroatian, Arabic,
Armenian, Chinese, Finnish, Kurdish and Swedish. We get along well although we come from different countries and
have different religions.
I am a Turkish citizen. I am thirteen years old and came here when I was seven, but some of my friends were only a few
months old when they arrived. Songul, for example, was two months old when she came from Turkey. Deniz was even
born here, but her parents are Turkish. Her cousin Kürsat was only two months old when he arrived. Jacklin came here
from Lebanon because of the war. Tsz Kai came here from Hong Kong. Patrik's parents are from Finland but he was born
here. Kati is from Poland. Sandra is from Yugoslavia like Dragan. Caroline is from Lebanon but speaks Armenian.
Magnus's mother

< previous page page_129 next page >


< previous page page_130 next page >
Page 130
is from the West Indies, but his father is from Sweden. Ann-Marie is from Skåne. 1 Caroline is from Chile like Pato and
Alejandro. Micke and Roland are Swedes. Rahime comes from a village in Turkey where they speak only Kurdish.
This is class 7D in the Flemingsberg School.
Aylin, 14

The Situation
In many suburban areas of the big cities in Sweden today there are a great number of immigrants, most of them having arrived
during the last 20 years. Contrary to what is usual in many other European countries, the immigrants do not live separated from
each other in ethnic groups. In one of the suburbs of Stockholm, Rinkeby, for example, about 100 languages are spoken and
more than 30 languages are taught at school. In many classrooms and kindergartens in these areas children representing first or
second generation immigrants are in an overwhelming majority. It is not unusual that in a class of 30 children only two or three
have a 'pure' Swedish background and that 10-15 different languages are spoken in the homes of the rest of the class. Some of
these children have arrived in Sweden relatively recently and have not yet mastered the Swedish language very well. Others
were born in Sweden, but use mainly one (or more) of the total of about 150 foreign languages spoken in Sweden today. In
some cases the immigrant children use their mother tongue in school, i.e. if they attend a so-called 'home-language class', and
those who attend a 'Swedish' class are entitled to at least two sessions a week of 'home language', if they wish. As for Swedish,
newly arrived children are given preparatory courses, and those who have spent a long time in Sweden but still have problems
with the language are given supportive courses. Most immigrant children, thus, can be claimed to be bilingual, at least to some
extent.
Children in those areas where many languages are spoken seem to use mainly Swedish as a lingua franca in contact with peers
(Boyd, 1985), irrespective of what language is used at home. In the Swedish debate, however, their mastery of Swedish
sometimes has been characterised as 'poor', and the children have been claimed to be 'semi-lingual', i.e. 'not fully competent in
any language'. It is also sometimes reported by teachers, social workers and others that these children, even those who have
lived in Sweden since their early childhood, speak 'oddly' and that their Swedish is at times hard for a native Swede to
understand, especially when they are speaking informally within the group, even though the children themselves seem to
understand each other perfectly well.
In the light of these facts I shall discuss two hypotheses regarding what might happen to Swedish as used by immigrant children
and to Swedish as a whole: (1) Swedish, as used by immigrant children, may show certain

< previous page page_130 next page >


< previous page page_131 next page >
Page 131
features, related to a creolisation process, and (2) the Swedish language may in the future show signs of influence from the
varieties of Swedish used by persons with immigrant background.
The term 'immigrant children' will in the following refer to children and teenagers living permanently in Sweden in the type of
areas described above, whose parents have immigrated before the birth of the child or who, themselves, have immigrated at an
early age, i.e. pre-school-age.
Immigrant Children's Swedish
In spite of the fact that, during the last two decades, immigrant children in rather large numbers have been growing up in
Sweden, few linguistic studies have been made on what kind of Swedish they in fact use. As far as we know, any one of the
following three descriptions might be valid:
1. The Swedish used by immigrant children does not in any important way differ from Swedish as used by Swedish children of
the same age and social class, raised in a monolingual Swedish environment.
2. As a consequence of incomplete learning, especially at the lexical level, immigrant children's Swedish is 'poor', i.e. it has
deficiencies, that make it unfit to express the speaker's elementary linguistic needs.
3. Immigrant children's Swedish fulfils vital linguistic needs, but it deviates in some instances from native Swedish and
therefore it is conceived of by native Swedes as 'odd', 'different', 'difficult to understand', etc.

In fact, it seems quite possible that immigrant children's Swedish may be described in all three ways. Some individuals may use
an idiomatic Swedish, perhaps sometimes even hypercorrect due to the fact that many of them have learnt Swedish at school as
a second language. Other children might well in certain respects have a 'poor' command of Swedish, which especially is
handicapping to them at school, while again other individuals might use a variety that does not impair communication but still
contains some deviant, and from a Swedish point of view, 'strange' elements.
It might also, however, be possible that the Swedish of one and the same individual could be characterised in all three ways,
different in different situational contexts. If so, his language in certain informal situations might seem quite 'Swedish' and typical
for his age, while in other situations, for example at school, it might be considered 'poor' and insufficient. Finally, the same
individual may, in a context consisting of those immigrant children from different ethnic groups, with which he normally
interacts, have developed a certain variety of Swedish, which could have deviant features. This variety might serve as a group
dialect with the purpose of identifying the members of the group as 'foreigners', a word that need not have any negative
connotations to the children themselves. This special immigrant sociolect might, if it exists (at least locally), serve as a group
ethnic variety

< previous page page_131 next page >


< previous page page_132 next page >
Page 132
in about the same way as Black English in the United States (Labov, 1972), and might perhaps be compared to other group
varieties like slang or secret languages.
The 'odd' deviations reported might be interpreted either as consequences of incomplete learning, resulting in incorrect inflection
morphology, incorrect word order, a limited vocabulary etc., or as the results of interference, i.e. the speaker transfers
grammatical rules and semantic features from his mother tongue when speaking Swedish. On the other hand, errors resulting
from incomplete learning do not normally obstruct communication to any large extent, unless they appear in a great number,
especially not if the listener (teachers, social workers and so on) has a long experience of talking to immigrants. This kind of
language is usually commented on as having 'plenty of errors' rather than as being 'odd'. We also know that transfer is not as
important in language learning as we thought earlier. Therefore it seems possible that those features which seem odd and
sometimes even unintelligible may have other explanations.
Consequently, it seems worthwhile to investigate immigrant children's language, not only from the perspectives of 'incomplete
learning' and interference but also from a point of departure, connected to our knowledge of what usually happens at instances of
language contact, creolisation and language change, at the same time taking into consideration sociolinguistic and
psycholinguistic aspects of language use.

Immigrant Children's Swedish and Creolisation


Pidginisation and creolisation
In those areas of the world where pidgins have originated, it has often been the case that people with different mother tongues
have lived together (sugar plantations in Hawaii, recently developed big cities in Africa and so on). The rising generations in
such areas usually learn one or two languages at home; for example the mother's and/or the father's mother tongue, and also the
pidgin used as a lingua franca by the adults in the neighbourhood (Spencer, 1971). This pidgin is normally the language used by
the children for communication with peers in the playground and so on. In some cases the pidgin has developed into a more
fully fledged language, a creole, that differs considerably from the base language, from which the pidgin once developed. Such a
creolisation process may take place very fast, sometimes within a few generations (Bickerton, 1981).
A situation that in many respects is similar to this 'classic' situation for the occurrence of a creole is at hand in Sweden today. As
we have seen, the linguistic situation in certain suburban areas is diversified. In addition to all the languages spoken by first
generation immigrants, such as Greek, Turkish, Spanish, Chinese and so forth, 'broken' Swedish is used as a

< previous page page_132 next page >


< previous page page_133 next page >
Page 133
means of communication between adults with different mother tongues. The Swedish, as heard by the children from staff in
shops and cafeterias, in the homes of their mates, and even at the kindergarten, where the members of the staff may originate
from different countries, is more or less influenced by foreign substrata and sometimes rather simplified. In the increasingly
common mixed marriages between two immigrants with different mother tongues, Swedish is mostly used as the language of the
family, but also here the variety may be quite different from native Swedish. For many children, school is the only place where
they hear Swedish spoken by natives. On the other hand, the Swedish pupils may be a small minority in the class, as described
in the composition above, and the teacher, having children with many different mother tongues with different mastery of
Swedish, may have to use a simplified variety of Swedish, best characterised as a combination of Teacher talk and Foreigner
talk.
It seems clear that the linguistic input, received by the children in areas like these, is quite different from the one received by
children in areas with few immigrants. One may therefore have reason to expect a different linguistic outcome from children
raised in districts where many immigrants live than from 'ordinary' Swedish children.
Is there a Swedish pidgin?
In a number of earlier studies (Kotsinas, 1981, 1982, 1984a,b, 1988a and others), I have dealt with the interlanguage used by six
adult immigrants in Sweden, five Greeks and a Polish woman. The six informants were unskilled workers, most of them had
limited education and none had attended any courses in Swedish, but had learned some Swedish during their 3-15 years in the
country.
The language that appears in these studies shows apparent similarities at all linguistic levels to the interlanguage used by
immigrants with other mother tongues (Finns, Arabs, Turks, etc.), and also to the interlanguage described in German studies of
Spanish, Italian or Turkish 'guest workers'.
Even more interesting, however, are the similarities between the six informants' interlanguage and pidgin languages in different
parts of the world, for example Hawaii, the Seychelles, and New Guinea (Tok Pisin) (see Kotsinas, 1985, 1987, 1988a).
Like most pidgins, the immigrant Swedish variety is characterised by a very limited vocabulary, omission of certain function
words, a reduction of the number of prepositions, almost complete absence of inflectional endings, few or no subordinate
clauses, deviations regarding word order, and a considerable variation among the speakers' varieties.
On the other hand, there are strategies to compensate both for the shortages in vocabulary and for the lack of grammatical
morphemes, such as semantic over-extensions, repetitions, circumlocutions and paraphrases. As for grammar, there also seems
to be the same tendency in Swedish

< previous page page_133 next page >


< previous page page_134 next page >
Page 134
interlanguage as in pidgins to prefer analytic, i.e. lexical, expressions in cases where the target or base language uses synthetic,
i.e. morphological, means of expression (cf. Traugott, 1973; Kay and Sankoff, 1974). In cases where the target language uses
synonymous expressions, one analytic (for example more fine) and one synthetic (for example finer) the learner or pidgin-
speaker seems to prefer the analytic one (Kotsinas 1985: 112), probably because an analytic expression is easier to apprehend at
the initial stages of learning.
Grammatical markers are also, as in pidgins, to some extent replaced by certain lexical material. Tense is for example marked by
time adverbs (i går (yesterday), i dag (today), i morgon (tomorrow) instead of by inflectional endings, and plurality by numerals
(två flicka (two girl)). Imperative and modality are often marked by words meaning 'please' and 'maybe' and there are tendencies
to mark aspect, which in Swedish, as in most languages, is a difficult category for learners, by using words meaning 'come', 'go',
'stay' and 'stop' (Kotsinas, 1988a).
Is there a Swedish pidgin, then? To answer this question we would have to know whether a restricted language of the type
described above is generally used as a means of communication in a certain area. This we do not know, since no sociolinguistic
studies of this sort have been carried out. All we know is that some of the prerequisites seem to be at hand, i.e. that at least
some adult learners of Swedish use an interlanguage that has many features in common with pidgin languages in different parts
of the world.
Creolisation
When a pidgin creolises it is no longer a more or less occasional contact language, used in a limited number of situations (trade,
conversations with superiors and so on). Instead, it is used more continually by a certain group of speakers, for example new
generations. In those situations it is faced with the demand of fulfilling all the functions that languages usually have in
communication. The many analytic means of expression in the pidgin variety, the frequent repetitions and semantic over-
extensions and the lack of morphological means seem tiresome and uneconomic, and as a consequence the pidgin may develop
in one of two possible directions. Either the speakers choose to learn the target/base language with all its irregularities and
opaque means of expression, or to create a 'new' grammar by developing certain lexical items, which in the pidgin variety have
replaced grammatical markers, into pure function words or affixes. Words meaning, for example, 'stay' maybe used to create
copulas, and words like suppose may, as in New Guinea Pidgin, be grammaticised to serve as subjunctions (Kotsinas, 1984c,
1987, 1988a; cf. also Sankoff and Brown, 1976). Which alternative the speakers 'choose' is rarely conscious, but related to
factors like degree of distance from and attitudes towards the society of base/target language speakers.

< previous page page_134 next page >


< previous page page_135 next page >
Page 135
Most people who learn a language without formal instruction use pidginlike varieties at the initial stages of the learning. Just
like pidgins, such a variety may develop in two directions. Either the immigrant learns to use all the essential parts of the target
language system, or he develops his variety along paths that are typical for early stages of creolisation. In the latter case the
speaker replaces grammatical means by lexical material, relies on certain compensatory strategies and so on, as described above.
Among the six speakers in my material one can distinguish two groups. Three of the speakers can be characterised as 'learners'
and the other three as 'pidgin speakers' (Kotsinas, 1985), and in studies of German Gastarbeiterdeutsch similar groups have been
distinguished (Meisel et al., 1981).
As for the second generation immigrants, the children born and raised in Sweden, similar paths may be possible. Either they will
learn idiomatic Swedish, or they will develop their own varieties of Swedish, based on the pidgin-like varieties that have formed
so much of their linguistic input.
A Swedish creole?
Is it possible then, that a creole will emerge in Sweden?
The emergence of a creole presupposes a distance, socially, psychologically and/or geographically, to speakers of the base
language. To a certain extent these kinds of barriers exist between native Swedes and immigrants, at least in the type of areas
described above. On the other hand, there are counteracting factors, like school education, migration within the country
(immigrants moving to areas with few immigrants), radio, television and so on. Therefore, it seems unlikely that a new language
in the form of a creole will develop.
This, however, does not mean that all immigrant children, not even those born in Sweden, should be expected to use a
completely native-like variety. It might well be that their language bears traces of the pidgin-like input as well as of constant
influence from the interlanguage of more recently arrived peers.
If so, one would expect a variety with lexical, i.e. analytic, expressions where Swedish children prefer morphological means,
simplifications within the opaque, i.e. difficult, parts of the grammar, a certain number of repetitions and paraphrases, and
perhaps also, as in creoles, additions to the vocabulary; for example, by loans and translations of lexical items from the
languages spoken in the area, all used without hesitation, just like an 'ordinary' language. Such a variety might also be very
fluent.
New dialects?
As a matter of fact there are signs indicating that varieties of the type described are already on their way (Kotsinas, 1988b). In
the above-mentioned suburb, Rinkeby, the teenagers themselves are very well aware of the fact that they speak differently from
children in other parts of the Stockholm area. They even have names for the variety, Rinkebysvenska 'Rinkeby-

< previous page page_135 next page >


< previous page page_136 next page >
Page 136
Swedish' or Rinkebyska 'Rinkebyish'. Asked about the characteristics of this variety, they answer that it is 'different' and 'more
tougher' and that it is 'filled with slang' and not used in the presence of adults. Some even mention that it is 'secret'.
The variety seems to contain simplifying features, mostly grammatical, but also elaborating ones, mostly lexical.
Most striking at first glance is the pronunciation. Both vowel quality and quantity seem to deviate from standard Swedish, and
the distinction 'long-short syllable' seems to be diminished or erased. On the other hand, certain phonemes are more distinctly
pronounced than normal. The prosody contains both 'Swedish' and deviant features. Remarkably, it is almost impossible for a
native Swede to guess the speaker's mother tongue, i.e. the accents seem to have converged into one, signalling 'Swede with an
immigrant background'.
The same mixture of 'Swedish' and 'foreign' characterises vocabulary, where slang words and other items, typical for teenagers,
occur together with words, used as slang, from languages spoken in the area, such as Romani choravles (thief) and ayde len, a
Greek and a Turkish word meaning together 'get lost'. Certain Swedish words and idioms sometimes have another semantic
content than normal, especially words meaning 'go', 'come' and so forth, which also in interlanguage undergo semantic extension.
From a Swedish point of view there are frequent grammatical errors, especially as regards rules for gender, agreement, word
order and the use of prepositions. Certain deviations seem so fixed that even some Swedish children and, occasionally, even
teachers and youth centre staff use them. So, for example, a boy, corrected by his teacher in a matter of grammar, once
answered: 'What you say is not correct in Rinkebyish'. A feature, common to almost all deviations, is that they show a direction
towards less marked structures, compared to the more marked Swedish ones, and, irrespective of the mother tongue of the
speaker, these deviations tend to be very similar.
So far this dialect in the making shows individual variations and is used by young people only, and in more pregnant form by
some individuals, mostly immigrant boys, than others. Quite in accordance with the speakers' own statements, it seems to be
used more frequently in certain contexts, for example at the youth centre, than in others, for example in classrooms. Whether it
is going to stabilise and develop or die out depends on a lot of different factors, such as education, migration and social
mobility, and is impossible to predict.

Language, Learning and Language Change


In a further perspective, however, there is still another interesting question to ponder upon, namely whether the development of
the Swedish language will be affected by the immigration. In other words, is it possible that, when

< previous page page_136 next page >


< previous page page_137 next page >
Page 137
in a few years a third of all children in Sweden will be of immigrant descent, this will not leave any traces in the language? To
discuss this question one has to consider both linguistic and sociological aspects.
Linguistic aspects
It is a well-known phenomenon that languages may undergo certain changes in a situation where two or more languages are in
contact.
It seems to be a prevalent assumption that, in cases of language contact, lexicon is easily affected by borrowings, while grammar
is hardly affected at all, with the exception of a few known cases of 'borrowings' of grammatical morphemes. Weinreich (1970:
113), however, remarks that changes may affect grammar as well and refers to the Balkan peninsula as a 'linguistic whirlpool'.
Gumpertz and Wilson (1971) have described deep-going grammatical changes as a result of long-term contact between four
languages in an Indian village. The languages are lexically distinct in almost every respect, but reductions and adaptations in the
direction of unmarked linguistic structures, regarding, for example, gender, copula constructions and subordination, have
resulted in word-for-word translatable, creole-like varieties. The cause of these changes is said to be that the constant code-
switching in daily life has to be done with a minimum of effort and additional learning.
Others have stressed input as important for language change. As for first language learners, Traugott (1977) has pointed out that
input may lead to language change, in cases where it is incomplete. Like second language learners, first language learners
perceive lexical (analytic) expressions for grammatical categories more readily than morphological (synthetic) ones (cf. Slobin,
1975). If a certain grammatical category, as a consequence of phonological changes, has become reduced and opaque, the child
tends to replace it with more salient, analytic ones. This may explain why the subjunctive form of the verb in Swedish, as in
English, after having coincided almost completely with the indicative, has been replaced by adverbs and auxiliaries which
express modality. Janson (1979) is of the opinion that Latin, in spite of its culturally dominating position, underwent certain
phonological, morphological and syntactic changes, when the number of first generation speakers increased substantially during
the first century A.D. as a consequence of political expansion. According to Janson these children had difficulties in perceiving
fine phonetic details and unstressed grammatical morphemes, which in turn resulted in a change-over to lexical markers.
In a situation like the one at hand, both input and the fact that the immigrant children in Sweden live in multilingual societies
with frequent code-shifts might contribute to the development of analytic and less marked expressions. It is generally
considered, however, that a language change in a contact situation cannot occur unless the native speakers accept it, and also
that such a change has to affect vulnerable parts of the language already on the way to change.

< previous page page_137 next page >


< previous page page_138 next page >
Page 138
Which parts of Swedish are so vulnerable, then, that a latent change may be accelerated? As we have seen, the most obvious
changes in Rinkeby-Swedish concern gender, agreement, prepositions, word order and the vowel system. These subsystems are
typologically very marked. The gender system, for example, is quite unpredictable, the prepositions are exceedingly many and
difficult even for native Swedes, and the vowel system can be said to contain 18 vowels. We know that marked systems are
difficult for language learners, irrespective of the fact that the learner may have a similar construction in his own language
(Hyltenstam, 1984), and it seems plausible that such systems are vulnerable also in cases of language contact.
Sociolinguistic aspects
From sociolinguistics we have learnt that linguistic changes often proceed from powerful groups to less powerful, and it may be
said that immigrant groups in Sweden today are on the outskirts of society, and that they, on the whole, are socially and
economically neglected and consequently would not be able to affect the development of the Swedish language. This, however,
is a somewhat simplified view. First, because the immigrants constitute a numerically large group, and, second, because they are
socially and economically active. Many smaller restaurants, shops and the like are run by first generation immigrants, who work
hard for the future of their children. The fact that many of them of their own free will have taken the step to leave their countries
in order to build a better future for themselves and their families is a sign of enterprise and daring, and it does not seem unlikely
that these qualities to some extent will be transferred to their children. It is, for example, a fact that immigrant children from
certain language groups exhibit a very high inclination for advanced studies (Löfgren, 1985). One might assume that at least
some persons with an immigrant background will play important roles in the future, economically, politically, culturally.
Another important aspect is that languages in fact do not change exclusively from 'above' (cf. Romaine, 1982).

Conclusions
Many children in certain areas of Sweden today get an input consisting partly of one or two 'mother tongues' other than Swedish
and partly of a simplified variety of Swedish. It has been suggested that deviations from standard Swedish should not be looked
upon only as 'errors' and the results of imperfect learning, but as features similar to those which appear in creoles. Therefore, at
least some children with an immigrant background might, if one looks at it negatively, have difficulties in learning grammatical
and stylistic nuances in Swedish. On the other hand, they might, if one looks at it positively, also have a greater capacity to
create new expressions for their linguistic needs.

< previous page page_138 next page >


< previous page page_139 next page >
Page 139
Hypothetically, it has been suggested that as a consequence of this, new dialects may emerge and that, in future, the Swedish
language might be affected. Whether or not changes like these will take place obviously has to do with not only linguistic but
also sociological and psychological facts.
Meanwhile, a situation like the one described in this paper certainly entails important problems for society and, not least, for
schools. The difficult question of which deviations in the language of children are signs of language change and which
deviations are just 'errors' is always present for the teacher, and even more so if the pupils are of more or less foreign descent
and the situation is as dynamic as it is at the moment in certain areas of Sweden.
The linguistic results of the huge immigration in Sweden during the last decades may not be seen, or even imagined, by us who
live in the middle of the process. Nevertheless, it seems important to investigate what goes on in front of our eyes. Doing so, we
might be able to exhibit a better understanding of what is going on in the language of the children, and also, in a further
perspective, provide an opportunity for posterity to watch new regional or social dialects emerge and to seize language history
on the wing.

Notes
1. Southern part of Sweden.

References
Bickerton, D. (1981) Roots of Language. Ann Arbor: Karoma.
Boyd, S. (1985) Language Survival: A Study of Language Contact, Language Shift and Language Choice in Sweden. Göteborg:
Department of Linguistics, University of Göteborg.
Gumpertz, J. and Wilson R. (1971) Convergence and creolization. In D. Hymes (ed.), Pidginization and Creolization of
Languages. Cambridge: Cambridge University Press.
Hyltenstam, K. (1984) The use of typological markedness as predictor in second language acquisition: The case of pronominal
copies in relative clauses. In R. Andersen (ed.), Second Languages: A Cross-Linguistic Perspective. Rowley, Mass.: Newbury
House.
Janson, T. (1979) Mechanisms of Language Change in Latin. Stockholm: Almqvist & Wiksell International.
Kay, P. and Sankoff, G. (1974) A language universals approach to pidgins and creoles. In D. DeCamp & J. F. Hancock. (eds),
Pidgins and Creoles. Current Trends and Prospects. Washington DC: Georgetown University Press, 61-72.
Kotsinas, U. -B. (1981) Come and Go: tense and aspect in immigrant Swedish. In N. Dittmar and P. Königer (eds), Proceedings
of the Second Scandinavian-German Symposium on the Language of Immigrant Workers and Their Children in West-Berlin,
Linguistische Arbeiten und Berichte, Heft 16. Berlin: Fachbereich Germanistik der Freien Universität, 112-29.
(1982) Repetitions in immigrant Swedish. In T. Bhatia and W. Richie (eds), Progression in Second Language Acquisition.
Special Issue of Indian Journal of Applied Linguistics, 9, 69-88.
(1984a) 'Ask maybe ten hours', Semantic over-extension and lexical over-use. Scandinavian Working Papers on Bilingualism, 2.
Stockholm: Department of Linguistics, University of Stockholm.
(1984b) On the acquisition of vocabulary in Immigrant Swedish. In H. Ringborn, (ed.), Förhandlingar vid Conference on
Psycholinguistics and Foreign Language Learning. Publi-

< previous page page_139 next page >


< previous page page_140 next page >
Page 140
cations of the Research Institute of the Åbo Akademi Foundation, 86. Åbo: Åbo akademi, 75-100.
(1984c) Om prepositionsanvändningen i nägra invandrares svenska talspråk. In K. Hylten-stam and K. Maandi (eds), Nordens
språk som målspråk. Forskning och undervisning. Stockholm: Department of Linguistics, University of Stockholm, 188-203.
(1985) Invandrare talar svenska. Malmö: Liber.
(1987) 'Varsågod inte problem'. Lexikala enheters grammatiska funktion i interimspråk. In E. Wande, J. Anward, B. Nordberg,
L. Steensland & M. Thelander (eds), Aspects of Multilingualism, Proceedings from the Fourth Nordic Symposium on
Bilingualism, 1984. Uppsala: Acta Universitatis Uppsaliensis, 401-18.
(1988a) Come, stay, finish. On the development of aspect markers in interlanguage and pidgin/creole languages. In L.-G.
Larson (ed.), Proceedings from the Second Scandinavian Symposium on Aspectology, Studia Uralica et Altaica, Acta
Universitatis Uppsaliensis 19. Uppsala: Almqvist & Wiksell.
(1988b) Rinkebysvenskan - finns den? In P. Linell, V. Adelswärd, P. A. Pettersson & T. Nilsson (eds), Svenskans beskrivnng
16, SIC. Linköping: Universitetet i Linköping.
Labov, W. (1972) The logic of nonstandard English. In P. Giglioli (ed.), Language and Social Context. London: Penguin
Education.
Löfgren, H. (1985) Elever med annat hemspråk än svenska. Pedagogiska uppsatser, 16. Lund: Pedagogiska inst., Lunds
Universitet.
Meisel, J., Clahsen, H. and Pienemann, M. (1981) On determining developmental stages in natural second language acquisition.
Studies in Second Language Acquisition, 3, 109-35.
Romaine, S. (1982) What is a speech community. In S. Romaine (ed.), Sociolinguistic Variation in Speech Communities.
London: Edward Arnold, 13-24.
Sankoff, G. and Brown, P. (1976) The origins of syntax in discourse: a case study of Tok Pisins relatives. Language, 52, 631-
66.
Slobin, D. (1975) The more it changes . . . On understanding language by watching it move through time. Papers and Reports
on Child Language Development, 10. Stanford University: Department of Linguistics.
Spencer, J. (ed.) (1971) The English Language in West Africa. London: Longman.
Traugott, E. Closs (1973) Some thoughts on natural syntactic processes. In C. Bailey and R. Shuy (eds), New Ways of Analyzing
Variation in English. Washington: Georgetown University Press, 313-22.
(1977) Pidginization, creolization and language change. In E. Valdman (ed.), Pidgin and Creole Linguistics. Bloomington &
London: Indiana University Press, 70-98.
Weinreich, U. (1970) Languages in Contact. The Hague: Mouton.

< previous page page_140 next page >


< previous page page_141 next page >
Page 141

Language Varieties and Intercultural Education


Sigrid Luchtenberg
FB 2, Essen University Universitätsstr.,
D-4300 Essen, West Germany.
Abstract. In most investigations of bilingualism as well as in the teaching of a second language, it is normally neglected
that LI and L2 each consist of a whole array of varieties with certain rules and functions depending on the situation of
communication, the role of the participants, etc. The present paper claims that language varieties represent an important
aspect of the development of bilingualism. This will be shown regarding the situation of migrant children in West
Germany. Thus language varieties have also to be taken into account with regard to language teaching (here: German as a
second language). Three kinds of competence will be expected of migrant children with respect to the different varieties:
productive, reactive and receptive competence. It will be shown that most of the language varietiesespecially situational
onesare connected with cultural developments. Therefore, they often show many sociocultural implications for language
and behaviour rules. Migrant children who fail in the understanding and use of language varieties often lack the necessary
sociocultural knowledge. Intercultural education turns out to be the method of instruction that could lead to better results.

Language Varieties and German as a Second Language


In language teaching we often neglect the fact that a language normally consists of a bundle of language varieties. In this paper
it will be shown that language varieties (like dialects, languages for special purposes, youth languages, etc.) have to be taken
into account when, for example, curricula for second language classes (such as GSL German as a second language) are planned
and learning difficulties of migrant children are discussed. My hypothesis is that such differences stem partly from sociocultural
differences and can therefore be solved in the context of intercultural education.
As the paper refers to the situation of migrant children in West Germany, their school conditions have to be described briefly.
Due to the Federal Constitution conditions differ amongst the 11 länder (cf. Röhr-Sendlmeier, 1986). Though these differences
concern many important questions, such as mother tongue teaching, compensatory GSL classes, etc., it can be stated

< previous page page_141 next page >


< previous page page_142 next page >
Page 142
that in general most migrant children go to so-called regular classes, i.e. German mother-tongue classes (together with German
students). Besides this, there exist several other types of classes which are attended only by non-German children, such as the
so-called 'bilingual classes' in Bavaria, homogeneous regular classes (i.e., only non-German students from one nationality) and
preparation classes for newcomers. Migrant students in regular mixed classes learn Germantheir second languagepartly in
additional GSL classes (normally 3-5 hours per week), partly in the regular German classes and partly 'naturally' in other classes
like science, and, of course, outside school.
Nowadays, many teachers complain that their migrant students fail in written texts (productive and receptive tasks), especially in
sciences and even in maths (e.g. text exercises) though they are fairly fluent in German conversation. Here, we have to ask
whether language varieties in which they have not been trained could be one of the reasons for these failures.
The goal of German as a second language is complete communicative competence (cf. Reich, 1976). This includes, of course,
language varieties, but we have to take into consideration that even native speakers usually do not know or practise all varieties
of their language. Therefore, it is difficult to decide which language varieties should be taught to non-German learners. Yet, as
the teaching of them has turned out to be necessary, the relevance of the different language varieties for migrant students needs
to be discovered with regard to school life and daily life situations.
Language varieties in German
Language varieties have mainly been studied for linguistic reasons. Due to changing research interests, some varieties have been
described often and carefully as, for instance, the varieties which are used by the military or those which are spoken by certain
professions like hunters, fishermen, etc. (cf. Nabrings, 1981). Language varieties are normally divided into temporal,
geographical, social and standard onesa categorisation that suffers from a lot of overlap, as can easily be shown when the use of
a dialect is considered to be a geographical, social and situational variety (cf. Fishman, 1979a).
Especially for didactic purposes it is more convenient to take situations as the starting-point and to inquire about the register that
is demanded in this particular situation. The register includes the adequate variety as well as certain behaviour rules, gestures,
non-verbal behaviour, etc. which are expected in certain communication situations (e.g. Gregory and Carroll, 1978).
Most migrant children have learned a geographical variety of German outside schoolsometimes even in school. This can be a
real 'dialect' or a geographically influenced everyday German. This also holds true in general for their families. Very often, they
are not aware of the fact that this German differs from the standard German. Among the adults there are migrants

< previous page page_142 next page >


< previous page page_143 next page >
Page 143
whose German consists mainly of the geographical variety of the region, mixed with some television German and some
professional language bits. Yet, these speakers do not know of these different origins.
In some parts of West Germanyespecially in the southchildren are allowed to use the geographical variety of the region in
kindergarten and in the first years of primary school. There, it is also known to their teachers that certain difficulties in
orthography (spelling), choice of words and even syntax are due to the differences between dialect and standard German for
German children (cf. Rascher, 1987). For migrant children, standard German sometimes has the character of a third language
besides their mother tongue and the local dialect, but the resulting difficulties have not been examined so far and most teachers
are not aware of them (e.g. Fishman, 1979b).
Social varieties consist of a bundle of different varieties which are spoken by social groups such as students, women, university
staff and motor cyclists. Most Germans belong to several of these groups and practise the special variety within each group.
Migrant students are, of course, also part of different social groups as, for example, their class, their school, their peer group,
perhaps their football club. They are confronted with professional languages, for instance, the language of instruction, maths,
sciences, etc. Due to personal habits, friends, etc., more social varieties can become relevant for migrant students.
Situational varieties are best described by Fishman's (1965) famous question: 'Who speaks what language to whom and when?'
They are determined by the entire communication context as, for example, the roles of the participants, the setting, the reason
for the interaction, the time, etc. The variety (as part of the register) that fits into a situation, depends on these components so
that, for instance, two seemingly similar situations, such as invitations for dinner, may demand very different registers (varieties
and behaviour rules); cf., for example, an invitation to a dinner party organised by students and an invitation to dinner by the
new dean of the faculty.
It is also important to stress that the conditions of communication situations (especially behaviour rules) change within societal
development. The large number of situational varieties makes instruction about them complicated, though there may exist some
general rules. Another important fact is the adequate behaviour that is demanded in communication situations. Oksaar (1981) has
shown that inappropriate behaviour because of situational interference leads more easily to misunderstandings and even rejection
by Germans than grammatical faults.
Language competences (L2) with regard to varieties
It has already been shown that not all varieties are spoken or even understood by all native speakers. The competences that are
expected differ with regard to the communication situations in which a person will be

< previous page page_143 next page >


< previous page page_144 next page >
Page 144
involved. Which language varieties are demanded by migrant students and should therefore be taught (or at least be explained)
in second language classes needs careful examination. We can differentiate three kinds of competence (e.g. Luchtenberg, 1988):
productive, i.e. to understand and to speak, perhaps also to read and write a variety;
reactive, i.e. to understand one particular variety and to answer in another;
receptive, i.e. to understand a variety.

Lack of competence normally leads to problems with regard to school and social life.
Certainly, productive competence is demanded by migrant children in the local dialect, a standard variety of German, and the
professional languages that are used in school. Students often use a group variety of their own (youth language) that must be
practised by those who want to belong to this peer-group. It is trivial to point out that lack of competence in the professional
languages and in standard German leads to school problems. Lack of competence in the local dialect may lead to an outsider
position (similar to the youth languages). Productive competence is also required in daily life situations, e.g. to buy milk in the
break or to answer an invitation to a birthday party.
Only reactive competence is required in the local dialect if the children move into this region at an older age. In school, the
language of instruction spoken by the teacher demands reactive competence, as does the literature that is read in class.
Deficiencies lead once again to school problems.
Receptive competence suffices, e.g. for television German, for the language of advertisement and partly for the language used in
newspapers. The latter may also demand reactive competence if newspapers are, for example, read in school.
Besides, it has to be pointed out that most communication situations require at least reactive competence in language and
behaviour.
Consideration has been restricted to L2 as language of the environment so far, with the result that competences of three kinds
are demanded for several language varieties, especially the local German, standard German and different professional varieties.

The Role of the Mother Tongue


The question of competence in language varieties has, of course, to include the mother tongue of the migrant children. Here, it
has to be taken into consideration that the mother tongue often does not correspond with the official standard variety, cf., for
example, Italian dialects. Sometimes the language in question is even from a different linguistic origin, like Kurdish as far as
Turkish children are concerned. This also holds true for different Yugoslavian languages.

< previous page page_144 next page >


< previous page page_145 next page >
Page 145
Usually, we expect the mother tongue (as the language of early childhood) to be developed within the family. It should normally
enable a child to participate in the interactions of his or her social group, e.g. including knowledge about the usual varieties and
registers. So, the role of the mother tongue for a bilingual child is mainly to be seen as an important component of social
development (cf. Belke, 1986). Beside this, many researchers claim that a fully developed and positively accepted mother
tongue will also help establish the second language. Influences from the mother tongue upon the second language are seen from
different viewpoints: interference as well as a positive basis for a second language (cf. Skutnabb-Kangas, 1982). Yet, it is
difficult to prove influences on language varieties because empirical studies are still lacking in this field. At least, there exist
studies of the situational interference that shows in communicative situations (cf. Oksaar, 1981; Schwenk, 1980).
Language varieties in L1 learned by migrant children living in a West European country will probably differ from those that
would be developed in the L1 country:
There are fewer occasions to listen to or read the standard variety with the exception of mother tongue classes, newspapers and
a few television/radio programmes and videos.
The everyday language has changed and will now include several Germanisms.

There aredepending on individual living conditionsfewer occasions to use a Turkish, Italian, etc., variety if, for instance,
shopping is done in German supermarkets, sports are practised in German clubs, and so on. Due to this, due to the reduced
family communication in families where both parents are working, due to the German school context, many topics are not, or
only seldom, discussed in L1. So, reduced capabilities in the use of language varieties in L1 have to be assumed.

Language Varieties and Bilingualism


Normally, a school beginner has developed the social variety of his or her group and will add at least the standard variety in the
very first years at school. According to this, a bilingual child would be expected to have competence in at least four varieties
(e.g. Extra, 1987 with regard to Turkish children in Holland): geographical-social dialect L1 and L2 together with standard L1
and L2. As a result, the goals of instruction would be to establish these four varieties within a bilingual education and to extend
the child's knowledge to further varieties, especially in German as the main language of instruction.
Yet, this can only be regarded as an ideal which seldom corresponds to reality in our schools. Besides the above-mentioned
reasons, several obstacles occur, amongst which the frequent lack of co-operation between German and mother tongue teachers
can be counted. This leads to the result that

< previous page page_145 next page >


< previous page page_146 next page >
Page 146
each of them judges a student to be fluent in the other language in which he or she hears a child speak, without being able to
recognise the deficiencies (e.g. Jørgensen, 1984 with regard to the Danish situation). Only a stable exchange of views between
teachers could change this situation. So, the above goal needs a careful approach, but as the demanded competences have
shown, it is necessary for a successful school life (cf. Philips, 1976 for an American example).
Diversity of language varieties belongs to communicative competence in L1 as well as in L2. Instead, we find migrant students
who lack this diversity in both languages. Possible explanations are:
L1: insufficient practice in the mother tongue; an insufficient number of situations to train situational varieties; an insufficient
number of occasions to listen to or to read a standard version of the L1; no, or not enough, encouragement to use and develop
the L1;
L2: insufficient practice in L2; no training in the switching of varieties; perhaps an insufficient number of situations to train
situational varieties; no, or not enough, explanations about and exercise in routine formulae and conventional patterns that are
part of many communicative situations;
L1 L2: no, or insufficient, teacher training in these areas; only a few contrastive studies in this field.

Learning Difficulties
One reason is to be seen in the fact that many teachers of GSL neither reflect nor teach the diversity of varieties, though they
expect their students to use them. Most GSL books neglect important structures (as, for example, passive voice) for a long time,
though they are needed to understand technical language even in primary school (cf. Luchtenberg, 1986). Besides other reasons,
there is a very important one for learning difficulties: varieties stem from a sociocultural background; they are full of societal,
social and cultural experiences which may be unknown to migrant children (cf. Oksaar, 1983; Scharnhorst and Steinmüller, 1985
for examples).
The cultural implications of language varieties are one of the most important reasons why migrant students (and adults, too) fail
so often in the correct use of varieties and registers in German. Of course, varieties, situationally adequate verbal and non-verbal
behaviour, phraseology, idioms, routine formulae, etc., are trained in kindergarten and school, but more or less from a German
background and viewpoint so that the needs of the migrant children are not met (e.g. Daniels and Pommerin, 1979). Therefore,
intercultural education seems to be the only possible solution to help develop full competence in language varieties.

< previous page page_146 next page >


< previous page page_147 next page >
Page 147
Migrant students are often judged in a wrong way because of sociocultural difficulties with regard to varieties and situational
behaviour. For example:
a child is unable to solve a task in science because of a lack of cultural knowledge, but the teacher notices only that the problem
has not been solved;
a child does not look into the teacher's eyes when he or she is addressed because of cultural traditions, but the teacher assumes
the child to be dishonest.

On the other hand, many teachers are mistaken in judging their migrant students to be fluent in German because they are fluent
in the regional variety in daily life conversations. It has also to be taken into account that the use of language varieties is closely
related to the question of norms that might be answered in a different way by different teachers.

The Role of Intercultural Education


With regard to language varieties and their sociocultural content, the role of intercultural education is mainly to be seen in the
fact that intercultural education wants to make the learner aware of cultural differences and similarities under the assumption of
equality. Intercultural education normally takes place in multi-ethnic classes where the needs, capacities and deficits of each
child are regarded as the basis of instruction.
Thus, intercultural education could create a climate where language varieties and situational behaviour can also be discussed,
explained and tried out (e.g. Hohmann, 1983; Pommerin, 1984). If co-operation with the mother tongue teacher is possible
(preferably team teaching), full bilingual competence in language varieties can be a realistic goal for migrant students, while
German students profit from language awareness and knowledge about cultural and societal backgrounds of linguistic and
behaviour rules.
It seems to be necessary not only to provide the chance to listen to varieties and to try them out, but also to give further support
by explaining idioms, by exercising typical structures and by facilitating the learning of pragmatic rules. This holds true for
German children as well, though many of them get further support from their families.
Many of the problems of migrant students stem from the fact that their knowledge is not accepted, because of different
sociocultural contents as, for instance, in sciences, German, arts, music. This holds especially true for the pragmatic rules
governing varieties and registers. Acceptance and discussion of different cultural attitudes within intercultural education are an
adequate solution that can enrich the experiences of all children. Possible ways of instructing and suitable situations are, for
example:
reading of migrant literature where sociocultural misunderstandings are often described (cf. Luchtenberg, 1987; Pommerin,
1984);
games, situational dialogues, role plays, simulations;

< previous page page_147 next page >


< previous page page_148 next page >
Page 148
topics (in sciences, etc.) with cultural implications like behaviour, food, family, shopping, illness, etc.;
reading and discussing of essays written by students (e.g. Hegele & Pommerin, 1983);
trips to places outside school (post office, station, hospital, etc.) to allow practical experiences.

Thus, children who begin to feel accepted can learn different verbal behaviours and varieties (besides other matters). Language
experiences shared by German and non-German children can help to overcome misunderstandings because all children learn
about situational interference and cultural implications. Furthermore, they learn to speak about language.
With regard to language varieties in a multi-ethnic classroom, it has to be asked whether there are differences from a
monolingual classroom and how to treat them. Certainly, there are differences to be heard:
all kinds of variations of German (i.e. learner languages on different levels as shown by Klein & Dittmar, 1979);
code mixtures (cf. Kotsinas, this issue, pp. 129-40, for a Swedish example);
mother tongues (perhaps in different varieties); and
even sometimes foreigner talk.

All of them should be explained and discussed: their function, their structures, their problems.
It should also be shown that most of them are also to be heard or read in our somewhat multilingual society. And
multilingualism is certainly an important theme in intercultural education (e.g. Gumperz, 1976). Cultural awareness of language
behaviour is a major aim of intercultural education and answers at the same time most questions of language varieties, especially
situational ones. Language behaviour includes a good deal of traditions, manners, etc.topics that are usually taught in language
classes abroad (though not always in an adequate way), but that have also to be discussed in classes in the L2 countryin
intercultural settings (cf. Holmes & Brown, 1984 for examples).
The necessity of teaching language varieties in school, because of their importance for migrant students, has been pointed out.
On the other hand, language varieties can also help to show that cultural and linguistic diversity has already existed within our
society so that multilingualism/multiculturalism means only a new dimension of an old fact.

References
Belke, G. (1986) Schulpolitische Voraussetzungen und sprachdidaktische Konsequenzen einer zweisprachigen Erziehung.
Diskussion Deutsch, 90, 379-88.
Daniels, K. and Pommerin, G. (1979) Die Rolle sprachlicher Schematismen im Deutschunterricht für ausländische Kinder. Die
Neueren Sprachen, 78, No. 6, 572-86.
Extra, Guus (1987) Zweitspracherwerb und Beschulungsmodelle unter vergleichenden Gesichtspunkten. Die Beispiele Holland
und USA. In E. Apeltauer (ed.), Gesteuerter Zweitspracherwerb. München: Hueber, 53-66.

< previous page page_148 next page >


< previous page page_149 next page >
Page 149
Fishman, J. (1965) Who speaks what language to whom and when? La Linguistique, 2, 67-88.
(1979a) Some basic sociolinguistic concepts. In H. T. Trueba and C. Barnett-Mizrahi (eds), Bilingual Multicultural Education
and the Professional. From Theory to Practice. Rowley, Mass.: Newbury House, 120-29.
(1979b) Standard versus dialect in bilingual education: an old problem in a new context. In H. T. Trueba and C. Barnett-Mizrahi
(eds), Bilingual Multicultural Education and the Professional. From Theory to Practice. Rowley, Mass.: Newbury House, 454-
66.
Gregory, M. and Carroll, S. (1978) Language and Situation. Language Varieties and Their Social Contexts. London: Routledge
& Kegan Paul.
Gumperz, J. J. (1976) Communication in multilingual societies. In F. Pialorski (ed.), Teaching the Bilingual. New Methods and
Old Traditions. Tucson: The University of Arizona Press, 100-6.
Hegele, I. and Pommerin, G. (1983) Gemeinsam Deutsch lernen. Interkulturelle Spracharheit mit ausländischen und deutschen
Schülern. Heidelberg: Quelle and Meyer.
Hohmann, M. (1983) Interkulturelle ErziehungVersuch einer Bestandsaufnahme. Ausländerkinder in Schule und Kindergarten,
4, 4-8.
Holmes, J. and Brown, D. F. (1984) Developing sociolinguistic competence in a second language. Teaching English as a Second
Language: Perspectives and Practices. Selected Papers. Cambridge, Mass.: Center for Bilingual Education, 141-51.
Jørgensen, J. N. (1984) Turkish - Danish. A dynamic-diachronic perspective on the linguistic contacts of a minority. Paper
presented to the Eighth Scandinavian Conference of Linguistics. In J. N. Jørgensen (ed.), Sociolinguistic Papers. Copenhagen:
Department of Danish, Royal Danish School of Educational Studies, 95-103.
Klein, W. and Dittmar, N. (1979) Developing Grammars. The Acquisition of German Syntax by Foreign Workers. Berlin:
Springer.
Luchtenberg, S. (1986) Fachsprache in der Primarstufe? Sachunterricht und Mathematik in der Primarstufe (SMP) 7, 257-63.
(1987) Migrant literature in intercultural education. Paper presented to the IAIE-Conference No. 4.
(1988) Varietäten des Deutschen und ihre Bedeutung für ausländische Deutschlerner und lernerinnen. To appear in: Zielsprache
Deutch, 1.
Nabrings, K. (1981) Sprachliche Varietäten. Tübingen: Narr.
Oksaar, E. (1981) Zweitspracherwerb als kulturelles Lernen. Ausländerkinder, 7, 58-66.
(1983) Fachsprachen, Interaktionale Kompetenz und Kulturkontakt. In H. P. Kelz (ed.), Fachsprache. Sprachanalyse und
Vermittlungsmethoden. Bonn: Dümmler, 30-45.
Philips, S. U. (1976) Acquisition of rules for appropriate speech usage. In F. Cordasco (ed.), Bilingual Schooling in the United
States. A Sourcebook for Educational Personnel. New York: McGraw Hill, 195-208.
Pommerin, G. (1984) 'Migrantenliteratur' und ihre Bedeutung für die interkulturelle Erziehung. HDF meint, 7, 8-13.
Rascher, R. (1987) Das Fränkische im Alltag, in der Schule und in den Medien. In E. Wagner (ed.), Das fränkische
Dialektbuch. München: Beck, 105-56.
Reich, H. H. (1976) Zum Unterricht in Deutsch als Fremdsprache. In: M. Hohmann (ed.), Unterricht mit ausländischen Kinder.
Düsseldorf: Schwann, 149-84.
Röhr-Sendlmeier, U. M. (1986) Die Bildungspolitik zum Unterricht für ausländische Kinder in der Bundesrepublik Deutschland
- Eine kritische Betrachtung der vergangenen 30 Jahre.Deutsch lernen, 1, 51-67.
Scharnhorst, U. and Steinmüller, U. (1985) Fachsprachen als Lehr- und Lernhindernisse im Unterricht mit ausländischen
Schülern. Info zur pädagogischen Arbeit mit ausländischen Kindern, 12, 57-78.
Schwenk, H. (1980) Türkisch - Deutsch. Kulturelle und sprachliche Unterschiede. Praxis Deutsch Sonderheft, 31-35.
Skutnabb-Kangas, T. (1982) Some prerequisites for learning the majority languagea comparison between different conditions.
Osnabrücker Beiträge zur Sprachtheorie (OBST) 22, 63-95.

< previous page page_149 next page >


< previous page page_ii next page >
< previous page page_151 next page >
Page 151

Supplementary Mother-Tongue Education and the Linguistic Development of Yugoslav Children in Denmark
Andrina Pavlinic-Wolf *
Institute for Migration and Nationalities Studies,
41001 Zagreb, Trnjanska b.b., P.O. Box 88, Yugoslavia.
Karmen Brcic*
Zagreb, Yugoslavia
and
Nadezda Jeftic1*
Copenhagen, Denmark
Abstract. After a section on supplementary mother tongue education for Yugoslav pupils speaking Serbo-Croatian and
some information on Yugoslav migration in Denmark, some preliminary results are presented of a sociolinguistic survey
conducted on 50 respondents aged 10-18, attending regular instruction and mother tongue classes in Copenhagen. The
research was carried out in eight European countries hosting migrant workers: Austria, Denmark, France, FR Germany,
Great Britain, the Netherlands, Sweden and Switzerland. The mother tongue of 342 respondents was tested by means of a
questionnaire (a total of 25 questions, five for each of the five linguistic levels). The children also wrote a composition
('My School', 'My Home', 'My Homeland Yugoslavia', or 'My New Homeland'). In addition to the linguistic methods of
gathering data on their written productions, some sociological and sociolinguistic parameters were applied to examine the
communication range of the mother tongue, e.g. the length of their residence and schooling in Denmark Yugoslavia, the
educational level of their parents, verbal interaction with family members' Yugoslav peers Danish peers (self-assessment)
and some other forms of mother tongue usage, the self-evaluation of mother tongue skills, the comparative assessment of
L1 and L2 skills, etc.

< previous page page_151 next page >


< previous page page_152 next page >
Page 152
The Teaching of the Mother Tongue and Culture: the General Situation of Yugoslav Children in and around Copenhagen
Some difficulties of supplementary education
In the school year 1985-86 mother tongue instruction in Copenhagen was organised for the speakers of 25 non-Danish
languages, in 300 classes, in co-operation with all the counties of the island of Sjælland. The school authority of the Copenhagen
Municipality is engaged in making practical preparations for the instruction through the Office for the children speaking a
foreign mother tongue and in employing teachers with appropriate qualifications and teaching experience from their native
countries. In the same school year, a total of 90 teachers and approximately 4,000 children, or more than three-quarters of pupils
from major immigration groups, were involved in mother tongue instruction. The implementation of mother tongue curricula is
in accordance with the Decree of the Ministry for Education of the 20th November, 1984. The above evidence is given in
Modersmåls undervisningen i Hovedstadsområdet (1985-86).
Reflected in the new 1984 regulations governing migrant education is an intercultural approach to the teaching of foreign
children and youth. However, the intercultural experience gained by some other immigration countries is not promptly accepted
and applied; Danish methods in educating migrant workers and their children are to be based on experimental work in classes
and on the scientific results thus obtained. Since 1982, Copenhagen has been experimenting with educational and vocational
guidance for foreign pupils in the last years of the Folkeskolen. The teachers' languages and cultures of origin play a crucial role
in guiding pupils and in contacts with their families (cf. Document MMG-3(87)17, Council of Europe, p. 11). It is
acknowledged in the regulations that their language is the means used by foreign children to gain knowledge of the culture and
traditions of their homeland and that it is the essential prerequisite for them to find it easier to manage in the country of their
parents' origin in case they make visits there. The learning of the mother tongue is given support because the home language of
the parents can be used by the children as an instrument helping to maintain the homogeneity of the family.
Danish society seems to be creating its own model for educating foreign workers and their children. Such efforts obviously have
to do with the Danish immigration policy aiming at a successful integration of the workers and their offspring into the host
community. Accordingly, Denmark does not accept 'temporariness' as a qualification of their stay, nor has she included in the
teaching of the mother tongue and original culture prep-

< previous page page_152 next page >


< previous page page_153 next page >
Page 153
aration of the children for return to their homeland. (On the other hand, the Yugoslav migration policy relies on the notion of
temporariness, and thence there is also a terminological difference: from Yugoslavia's standpoint, her citizens working in
European countries are most frequently referred to as (labour) migrants, not emigrants/immigrants.)
As regards the return of Yugoslav migrants and their children, there is every indication 2 that it is a strong desire, even an
obsession with the majority of first generation migrants despite, or maybe because of, the fact that the interests that made them
leave the country of origin have already been satisfied. With 'second generation' migrants, return from Denmark is a result of the
parents' patriarchal desire 'for the children to return to the land of their grandfathers' rather than their own decision.
With regard to mother tongue instruction, the children may sometimes be urged by the parents' motivation to take it up. On the
other hand, a minority of Yugoslav parents would not encourage the children's study of the mother tongue in the first 6-7 years
of their life, from fear of the two languages' interference at such an early age, and in a firm belief that Danish is more important
for them as a first language on account of their embarking on a school life in Denmark.
This is one of the problems facing teachers engaged in supplementary mother-tongue education, who must then 'start from
scratch' with the children who are several years behind the language development of their peers in Yugoslavia.
The fact that mother tongue instruction is optional, which may aggravate truancy and drop-out rates, is also a problem.
Yugoslav pupils do not seem to have difficulties with the Danish language, particularly if they were born in Denmark or
socialised there at an early age. Their parents, who are always busy earning kroner, used to take them to the vuggestue (nursery)
first, and to the børnehave (kindergarten) next; the children would finally spend their leisure time at the fritidshjem before being
taken home after regular working hours. In other words, ever since their early childhood they have been taken care of by Danish
educators, psychologists, etc. Such circumstances are, of course, unfavourable for their mother-tongue acquisition and
development.
The learning of the mother tongue and culture was first organised for Yugoslav children in Denmark in the year 1973 (according
to evidence in Informacija o nastavi materjegjezika i nacionalne kulture za jugoslovensku decu u inostranstvu, Beograd 1983).
Supplementary mother tongue education is provided within the national educational system, with the Danish authorities in
charge of the selection and supervision of (Yugoslav-born) teachers and of the (Yugoslav) curriculum and textbooks, as well as
of finances. As for the languages of Yugoslav nations and nationalities (i.e. national minorities) that are taught in and around
Copenhagen, there are Serbo-Croatian, Macedonian, Turkish for Yugoslavs and Albanian, and earlier there used to be

< previous page page_153 next page >


< previous page page_154 next page >
Page 154
Slovene. In Copenhagen alone there are as many as 18 Serbo-Croatian classes, including one for pre-school children. (Some
municipal governments in greater Copenhagen allow organised mother tongue instruction at the preschool level, but the more
conservative ones seem to be determinedly against it.) There are nine more Serbo-Croatian classes in Frederiksværk, Helsingør
and Hillerød. Serbo-Croatian is taught by seven teachers, Turkish for Yugoslavs by two (who are Macedonian teachers as well,
along with another one for Macedonian), and Albanian by two. (The evidence concerning the number of Serbo-Croatian classes
and of Yugoslav teachers engaged in supplementary mother tongue education comes from the above-cited Danish source.)
According to the information of the Embassy of SFR Yugoslavia in Copenhagen, there are approximately 1,000 Yugoslav
children and youths, and about 7,500 Yugoslav migrant workers with the remaining members of their families. (Schierup's
estimate (1986: 78) is very close: 'In Denmark immigrants from Turkey are the largest single immigrant group in the country
with nearly 20,000 and Yugoslavs the second with about 7,000.')
Supplementary mother tongue instruction is held once or twice a week, lasting a total of four schoolhours, mostly at a time when
the pupils are already tired after regular classes, or on Saturdays. It may clash with their sundry leisure-time activities (soccer,
sewing, etc.), which does not add to its popularity among Yugoslav children. The minimum number of pupils per class being 12,
children of different ages are often put together to form a class (e.g. first to third form pupils, fourth to sixth form ones, seventh
to tenth form ones; there are even classes holding first to tenth form pupils). This can hardly suit children, and their teachers find
it difficult to cope, too.
The duration of this type of education is 9 + 1 years (the final year is attended by the best only). The drop-out rate is very
pronounced after the eighth form (when a class is generally halved).
There is no specialisation of teachers by individual subjects, but instead each is responsible for them all: the mother tongue
(including singing, poetry, theatrical improvisations), history, science, geography, Marxism and social system, comparative
cultures, etc., which must be demanding. Yugoslav teachers are placed under an obligation to teach their pupils 27 hours a week.
At a teacher-training college in Copenhagen they can acquire the national teacher's qualifications, having mastered a four-year
course.
In addition to information, data and impressions concerning supplementary mother tongue education in and around Copenhagen,
some subjective and objective difficulties and problems accompanying Yugoslav classes have been hinted at. There is also a
specific one that Serbo-Croatian teachers are faced with, which deserves to be mentioned: some Yugoslav children attend
'mother tongue classes' although the medium of instruction and the target language is not at all their home language, but the
language of social setting in their homeland! Reference is made here to Wallachian pupils in the first place, who start acquiring
Serbo-Croatian when they embark on mother

< previous page page_154 next page >


< previous page page_155 next page >
Page 155
tongue instruction. (A teacher told us that every now and then some Wallachian child wetted the classroom because he or she
did not know how to ask permission to go to the toilet.) Cases in point are sometimes also Romany children, and members of
the Romanian nationality (national minority) in Yugoslavia.
There follows an account of the ethnic, social and sociolinguistic composition of Yugoslav migrant workers, speakers of Serbo-
Croatian, and of their immigration into Denmark, for the most interested readers to gain a better insight into the sociocultural
identity card of the migrant children, our respondents in Copenhagen.
The first Yugoslavs to come to Copenhagen arrived through employment service agents in the 1960's. Although mostly peasants
(and unskilled workers), they would not receive employment in Denmark's agriculture, but in the heavy industry (iron),
shipbuilding and the textile industry (Silkeborg). Theirs were usually elementary school qualifications (4-8 years' education).
They were later joined, through chain migration, by their dependants (the so-called family reunification process), relatives,
friends, neighbours.
Major concentrations of Yugoslavs can be found in and around Copenhagen: in Hillerød, Næstved and Frederiksværk, where the
population of originating from eastern Serbia predominates. In Helsingør, however, the Yugoslav community is a mixed one,
and there are Romanies, too. The island of Sjælland holds about 80-85% of Yugoslav migrants. Two towns outside Sjælland are
inhabited by most of the remaining 15-20%: Odense and Silkeborg. During the economic boom the latter received about 300
Yugoslav girls originating from eastern Serbia and Bosnia, who got jobs in the textile industry and started raising their families.
The most common sectors of Yugoslav workers' employment today are: manufacturing industry, the plastics industry, the
precision industry (gauging, regulation instruments) and, decreasingly so, shipbuilding (welding). Women generally do cleaning
jobs (in hospitals) and work in laundries. As there may be only about 30 'Danish' Yugoslavs with higher education (college or
university) skills, this country evidently hosts Yugoslav labour rather than brain-drain-type migration. According to the
information of SFR Yugoslavia's Embassy in Copenhagen, the Yugoslav migrant population in Denmark is characterised by a
more or less offset migratory balance or a slightly rising one, relatively few naturalisations (about ten per annum), and a small
number of mixed marriages.
With regard to their ethnic composition, the Yugoslav migrant population could be roughly divided into three groups: about one-
third are migrants originating from SR Macedonia, Albanians and Turks by national affiliation; a solid third are Wallachians
who come from the plains of the Negotinska Krajina region in north-eastern Serbia (from border regions between Yugoslavia
and Romania, and primarily two communes there, Negotin and Kladovo; according to the Yugoslav population census of 1981,
SR Serbia is inhabited by 79.8% of the Wallachian total in Yugoslavia); almost a third

< previous page page_155 next page >


< previous page page_156 next page >
Page 156
are Montenegrins from Ivangrad and the former inhabitants of SR Bosnia and Herzegovina (who mostly come from Zepce * and
Mrkonjic-Grad*); SR Croatia and SAP Vojvodina are under-represented in Denmark as Yugoslav labour-sending regions, with
about 200 Croats and 100 Vojvodinians from southern Banat (the Yugoslav Embassy's information).
The second group is very interesting indeed for anthropological, sociological, sociolinguistic, cultural and other studies, and
numerous Yugoslav and foreign scholars, including Scandinavians, have done research on the Vlasi (singluar: Vlah)-Wallachians
(as a valuable source of information consult, for instance, Schierup (1985). We are also going to allocate Wallachians some
space relying mostly on this particular source.)
Schierup (1985: 26, note 20) states that the first migrants from the Negotinska Krajina region went away to Scandinavia in
1964. Mass emigration of Wallachian peasants from the region started in 1969. The Wallachians, representing one of the oldest
ethnic groups in the Balkans that had first come to the area in the seventh and eighth centuries, chose to settle down in southern
Sweden and in eastern Denmark. Besides primeval Balkanic roots and early Roman influences, the most important cultural
heritage they have carried with them to Scandinavia is later Slavic influences, predominantly Serbian: (cf. Schierup, 1985: 5). In
Sweden and Denmark, where they reside in a number of provincial towns, close to one another, they form a relatively closely
integrated ethnic minority group, maintaining an intensive internal social and cultural life (Schierup, 1985: 3). Their exact
number is unknown, because they are usually registered as Serbs (by their own choice) in Yugoslav population censuses and
migration statistics. 'Swedish' Wallachians often go to Denmark, where they, together with 'Danish' Wallachians, celebrate
important Yugoslav holidays. The two Wallachian groups across the Danish-Swedish border also intermarry or 'import' spouses
from the country of origin (Schierup, 1985: 5, 6 and also Schierup, 1984).
Unlike the Serbs in the Negotinska Krajina region, Wallachians in the same region are mostly a rural population, with poor
education levels and low social mobility. In Scandinavia they seem to have adapted to the new environment, but we believe it is
thanks to their own compatriot networking in the first place.
Their home language is vlaski* (i.e. of a Wallachian), a Romanian dialect which is not their written standard. The dialect
abounds in Serbian loan words. In Yugoslav schools, Wallachian children attend Serbian classes only, because it is considered
by their parents to be of utmost importance for acquiring solid skills and good jobs. The same pattern is repeated in Scandinavia:
Wallachian migrant children study Serbian as their home language, partly because of a Serbian national self-identification by
their parents, and partly in order to be able to obtain jobs in case they return to their country of origin (cf. Schierup, 1985: 5).
Among children between 7 and 14 years of age born to Wallachian parents in Sweden and Denmark more than 50%

< previous page page_156 next page >


< previous page page_157 next page >
Page 157
are continuously residing abroad (Schierup, 1985: 26, note 23, and also Schierup & Ålund, 1985).
The regions of origin of the above-mentioned ethnic groups of Yugoslav migrants in Denmark are underdeveloped, mostly rural
regions in the four Yugoslav republics having Serbo-Croatian as the prevalent language of the social setting: SR Serbia with
SAP Vojvodina, SR Bosnia and Herzegovina, SR Montenegro, and SR Croatia (unlike the remaining republics, SR Slov enia
and SR Macedonia). In the emigration regions, as well as the rest of the four republics, standard Serbo-Croatian is taught at
school, and it is the language of the mass media and of formal usage. In informal situations, however, local dialects are spoken.
The dialects of Ivangrad and its surroundings, of Kladovo and Negotin and theirs, of Nis * and its neighbourhood, and of
southern Banat are characterised, among other things, by a great deal of morpho-syntactic simplification of the case system,
which is reflected in Serbo-Croatian in Denmark. The majority of Yugoslav pupils there encounter the standard usage of this
highly inflected language when starting their supplementary mother tongue education. Some of them do not seem to distinguish
between the two main varieties of the language: Serbo-Croatian ('eastern') and Croato-Serbian ('western') ones, of which
distinction there is awareness in Yugoslavia, but refer to it as jugoslovenski (SC) or jugoslavenski (CS) (i.e. Yugoslav) language.
In the course of our survey in Copenhagen, it was observed that the respondents took a great pride in their mother tongue, which
appeared quite distorted in some cases. They all made very good efforts to fill up the (socio)linguistic questionnaire and to write
a composition for the purposes of our research project, which was quite moving in the few instances (Wallachian children) when
they strove to achieve more than they actually could under the circumstances.

Research Project 'The Language of Yugoslav Children Living in European Countries'3


Introduction: the goals of the project
Depicted in this paper are only some preliminary results of a survey conducted in Copenhagen in the summer of 1986. The
survey was carried out in a total of eight European migrant-receiving countries, including 342 children of Yugoslav labour
migrants, speakers of the Serbo-Croatian language.
The purpose of this project launched by the Centre for Migration and Nationalities Studies in Zagreb and headed by Andrina
Pavlinic-Wolf* is to find out about the mother tongue (L1) of 'second generation' migrants living and receiving their education
in the industrial countries of Europe, to identify their language problems and isolate their causes. In focus are such

< previous page page_157 next page >


< previous page page_158 next page >
Page 158
sociolinguistic problems as seen to be shared by the majority of Yugoslav children and youths in the eight host countries, as well
as those that appear to be specific to respondents in individual countries.
The sample and the research methodology applied
For the sake of easier organisation, the survey was conducted on an equivalent sample of 50 Yugoslav children, speakers of
Serbo-Croatian, attending three different mother tongue classes (Yugoslav supplementary education) picked out at random. The
number of 50 entails a limitation on the sample's being representative, yet the data thus collected may be relevant and
representative of the group of migrant children comprising it.
In each of the host countries concerned, 45-50 children of Yugoslav origin were surveyed (the sample being somewhat smaller
only in Switzerland and in Great Britain), of both sexes, aged 10-18. The mother tongue of the respondents' of parents (whether
living together or separated) is Serbo-Croatian, too. The respondents should have spent at least four years of their lives in
foreign countries (a foreign country, by our definition, is any country but SFR Yugoslavia, although a respondent may have been
born in it).
The questionnaire consists of three parts. By means of the first one (52 questions) sociodemographic data on the children were
collected and attempts were made to find out, in an indirect way (i.e. through their self-assessment), about the communication
range of their mother tongue (our hypothesis: the greater the range, and provided the language performs more functions, the
closer it is to the standard in the country of origin and the less shift-prone). The second part is a multiple choice test, examining
the respondents' L1 competence at five linguistic levels (phonology, morphology, syntax, vocabulary, and orthography). The test
contains 25 questions (five for each level), with three choices each (i.e. only one correct, one incorrect, and one 'very incorrect').
The respondents were asked to pick out the correct answer. Finally, they were required to write a composition ('My School', 'My
Home', 'My Homeland Yugoslavia', or 'My New Homeland'). All three parts of the questionnaire would be put into correlation,
and the data thus gathered would be processed by means of appropriate statistical methods.

Preliminary Results of the Survey 4


Respondents in Denmark and their sociocultural background
The sample in Copenhagen consisted of 50 children of Yugoslav origin: 20 boys and 30 girls (N=50). Their average age: 13
years 4 months. They attend regular instruction in accordance with their age. Forty-three

< previous page page_158 next page >


< previous page page_159 next page >
Page 159
respondents or 86% 5 were born in Denmark, 7 in Yugoslavia. 29 respondents never lived in Yugoslavia, 9 spent up to three
years there, 6 five to seven years, and 6 nine to eleven years. The average stay in the country of origin of those respondents who
spent a fairly long period of time there (N=12) amounts to 7.8 years, and if we counted its duration on the basis of the total N
(=50), it would amount to 2.3 years. Concurrently with this, such an unfavourable prerequisite for learning and developing their
mother tongue is quite pronounced.
The respondents claimed (one did not reply) that 11 of their fathers did manual jobs, 21 did unskilled or semi-skilled jobs, 5
were craftsmen/artisans or skilled workers, 2 did highly skilled jobs, and as many as 10 were dependents.
Thirteen of the respondents' mothers do manual jobs, 17 do unskilled or semi-skilled jobs, 1 is in crafts/has a skilled job, 4 do
highly skilled jobs, and 15 are dependants.
Most of the working fathers and mothers perform jobs in accordance with their qualifications/skills: 7 fathers completed up to
four forms of elementary school, 11 had 4-8 forms of elementary school, 7 gained vocational or secondary school education,
and one went to college or acquired university education. One mother never went to school, 9 had up to fourth form of
elementary school, 13 completed 4-8 forms of elementary school, 3 attended schools for artisans/craftsmen, and 2 went to
college or university. Twenty-three children were unable to supply data on the education of their fathers, as were 22 children
regarding that of their mothers.
Forty-five respondents live together with both of their parents, 2 live with their fathers (without mothers), 3 with their mothers
(without fathers), 37 with their siblings.
Forty-eight respondents stated that they had been attending regular school in Denmark, while two claimed that they went to a
school reserved for foreign children. Only 16 children had some schooling in Yugoslavia (one year on average), the rest of them
have been attending school in Denmark. (Consequently, their mother tongue acquisition and skills have not been influenced
substantially by the Yugoslav school system.)
The school rating of Yugoslav pupils in Danish schools is very good indeed: 27 children are excellent pupils (5 of them spent 4-
6 years in Denmark, 15 spent 10-12 years, 5 spent 13-15 years, one spent under 3 years, and one spent 16-18 years), 20 are very
good pupils, only one is fair, and two children were undecided as to what to say. (The length of their stay in the host country is
evidently in correlation with their success at school.)
All the 50 respondents feel they have been accepted well by their Danish peers. As many as 36 of them attribute the fact to a
very good grasp of the majority language, four believe this may be so because they are very cordial and friendly with the Danes,
four think of themselves as highly communicative, two as adaptable to their environment, one is sought out

< previous page page_159 next page >


< previous page page_160 next page >
Page 160
'because his milieu needs company', three could not supply an answer. The respondents who made up our sample could
obviously not complain of xenophobia or any sort of negative feelings on the part of their Danish peers. (It is a different matter,
however, how much their feeling of belonging to the host community may interfere with a need for maintaining and developing
their mother tongue which, naturally, cannot be practised with their Danish friends.)
The same feeling of their being accepted as 'partners' has resulted in their relative non-selectivity in the choice of friends: 13 of
them have made friends equally with Danish and Yugoslav children. However, quite a number of them prefer mixing and
making friends with their compatriots, because it means maintaining links with their homeland in a way: 31 'to communicate in
the mother tongue/so as not to forget the language', four because they 'like their homeland', one because 'Yugoslavs do not look
down' upon him; one prefers Danish company 'to learn Danish'.
Starting from the premise that participation in leisure-time activities may be seen as one of the indicators of our respondents'
level of integration into Danish society, we have acquired some information in this connection, too. However, the data would be
more meaningful if they could be compared to a sample of our respondents' Danish peers. Namely, 24 of them are members of a
school organisation, association, society or club pertaining to sports, cultural, artistic, technical or similar activities, 25 are not.
(One child did not reply.)
A similar number, 23 respondents, do not frequent clubs or other venues for Yugoslavs, 24 seldom go to such places, only 3
attend often. Most of those who take an active part in the activities of Yugoslav clubs, 10 of them, recite poetry, 2 of them
engage in folk dances, one in 'something else'.
Unlike Yugoslav children in Sweden, all of whom stated that they had mother tongue instruction as a school subject, our
respondents in Denmark, 41 of them, claimed they had no Serbo-Croatian as a subject of instruction, while 8 said they did have
it and that they had been attending mother tongue classes. (According to all respondents, they do not have Slovene or
Macedonian as school subjects, either.)
To the question on through whom they had acquired most of their mother tongue, our respondents supplied the following
answers: 25 of them through their parents, 3 from siblings, relatives, friends, neighbours, 4 from Yugoslav teachers, prior to
departure for Denmark, 6 in their homeland, spontaneously, and 12 mentioned various other sources.
Nineteen children assessed their mother tongue skills as 'excellent', 19 as 'very good', 10 as 'good', and only two as 'bad'.
In addition to supplementary mother tongue education, mother-tongue-as-a-second-language classes, their own home, a
Yugoslav club or some similar organisation, the L1 standard can be made more accessible to migrant children in some other
ways, i.e. through visits to relatives and friends in the country of origin, organised tours, by means of books, newspapers and

< previous page page_160 next page >


< previous page page_161 next page >
Page 161
periodicals in the mother tongue, radio and television broadcasts, correspondence with compatriots, etc. Below is some
information on how much our respondents are engaged in such activities.
A great many, 42 of them, visit their homeland at least once a year, four twice a year, and another four 3, 4 or more times. The
length of stay varies, from 15-30 days (11) to 1-2 months (35), most probably during summer holidays. Few children (four) stay
longer than two months. Only 11 respondents have made organised trips to Yugoslavia for summer or winter holidays or for
excursions.
Thirty respondents have been in correspondence with someone in Yugoslavia, 27 in Serbo-Croatian, two in Danish, and one
failed to mention the language.
Our respondents' mother tongue skills are, obviously, the result of uneven and combined influences of various factors, as well as
of individual preoccupations with the language.
The respondents' L1 and L2: comparative self-assessment values
According to their comparative assessment of L1 and L2 skills, 24 respondents are equally good at L1 and L2, 6 are much better
at L1, and 20 are less good at L1. For the sake of its comparison with L1, below is some further information on our respondents'
L2.
The Danish language, according to 25, has been mostly learnt at school from local teachers, but 12 children think they owe their
L2 mastery to their friends, neighbours, acquaintances; six have learnt it from their parents, and one claims he learnt it in
Yugoslavia (which must be a mistake, because Danish is not one of the second or third languages to be taught in Yugoslav
schools); one respondent (who has spent less than three years in Copenhagen) admits that he has not yet acquired it, and five do
not know how they have mastered it (but they have). As many as 31 children believe their Danish is 'excellent', 14 'very good', 3
'fair', and 2 'poor'. Forty-four respondents had no difficulties with the language at the beginning, while six did (twotheir
communication was hampered; twoduring lessons; twoat play). Thirty-nine claim they are having no difficulties with the
language today (it would be logical to expect a greater number without any difficulties as against the preceding data, but the
respondents' L2 competence might be over-estimated), eight state a communication problem seldom comes up, and three feel
they have fewer language problems than when they first came. On the other hand, most of the children, 41 of them, help their
parents and compatriots in their interactions with their environment (i.e. when filling in forms, going shopping, etc.); eight do
not do that (and one gave no reply).
The frequency of L1 and L2 usage varies and depends on the occasion and on the interlocutor. The mother tongue is used, in the
first place, for communication with parentsin the case of 47 respondents (while two

< previous page page_161 next page >


< previous page page_162 next page >
Page 162
converse in Danish even with the parents, and one did not reply). In interaction with their siblings 22 children prefer their
mother tongue, but note that another 22 prefer Danish in the same situation. (6 children stated they had no siblings.) As many as
38 prefer Danish at play, 10 prefer their mother tongue, and 2 use both languages on an equal footing.
For 15 respondents the language in which they always think is Serbo-Croatian, for 25 it is mostly Danish, 2 translate mentally
from Danish even when speaking Serbo-Croatian, 2 think in the language in which they are addressed, and 6 could not supply a
reply on this question.
Some conclusions on the dominance of L2 may also be drawn on the basis of our question on the preferred language for the
purpose of writing. For 34 respondents it is Danish, for 14 it is their mother tongue, one child is equally good at both languages,
and one finds it easier to write in a third language.
Forty-four children read literature and periodicals in Serbo-Croatian, six do not; most of the readers belong to the group of
students who have spent 10-12 years (18) and 13-15 years (10) in Denmark. However, when asked whether they read more
frequently in Danish or in Serbo-Croatian, 28 respondents answered they preferred Danish, 19 claimed they made no difference
between the two, and only three stated their preference was their mother tongue.
A great number of our respondents in Sweden (N=50), 37 of them, follow radio and television broadcasts in Serbo-Croatian. In
Denmark, fewer respondents are interested: 25 follow such broadcasts, 24 do not (one did not supply an answer). Among the
most popular radio and television broadcasts are: Radio Belgrade's 'Horizont' (Horizon), radio/television news broadcast by RTV
centres in Belgrade or Zagreb, 'Pasos' * (Passport), 'Folk parada' ('Folk Parade'), 'Aktuellt for immigrants', 'Dobar dan,
Jugoslavijo' ('Hello, Yugoslavia!'), 'Språka', and 'Selo veselo' ('A Cheerful Village').
The results of testing L1 at five linguistic levels
As stated earlier, five times five questions were posed to represent each of the five linguistic levels. As N=50, the maximum
possible total of correct answers for each level is 250.
The majority of correct answers was obtained at the level of syntax: 228 or 91.2% (22 answers were incorrect). The respondents
were also successful at the levels of semantics and phonology, with 209 correct answers or 83.6% at each of these. (At the level
of syntax, however, there were fewer incorrect answers37than at the level of phonology39, but at the level of semantics 4
questions remained unanswered, and at the level of phonology just 2). Poorest scores were obtained at the levels of morphology
and orthography. At the former, there were 189 correct answers or 75.6% (57 incorrect ones, 4 questions remained unanswered),
at the latter 183 correct answers or 73.2% (67 incorrect ones).6

< previous page page_162 next page >


< previous page page_163 next page >
Page 163
On the whole, irrespective of the linguistic level concerned, the respondents gave correct answers in 78% of cases, incorrect
ones in 21.4%, and they abstained from answering in 0.6% of cases.
Some information on the compositions 7
Five of the respondents were unable, did not want, or did not have time (although they had plenty of time at their disposal) to
write a composition. The preferred topic for most of those who wrote one was 'My School' (15), and next 'My Homeland
Yugoslavia' (12). 'My Home' was the topic selected by ten pupils, and 'My New Homeland' by seven. A girl chose to write on a
'Hare', although the titles of the four given topics had been repeated to the respondents several times.8
For the purpose of assessing the respondents' compositions as objectively as possible, an encoding plan has been worked out for
registering a number of language variables (i.e. composition length, person, gender, number and case congruence, case syntax,
verb conjugations, aspect of verbs, c/c* distinction, capitalisation, punctuation, etc.) This segment of results will not be presented
here as it would be of limited (Serbo-Croatianists') interest. Suffice it to make some general remarks on the compositions and to
list several examples of Danish interference in the mother tongue of 50 Yugoslav children in Denmark.
The conspicuous characteristic of the compositions is their shortness: one wonders whether to call (11 respondents') 1-5 short
sentences a composition. There are also compositions one perceives as stream-of-consciousness sequences, with hardly any
punctuation marks or sentence markers, while the rest of them generally contain very short and simple sentences. It would be
difficult to make an intelligent guess at a respondent's age on the basis of his or her composition, as most of the children have
attained the same literacy level. On the other hand, several respondents have manifested some kind of garrulousness; despite a
very limited syntax, they must have felt an urge to pour out their stories, using the spoken, rather than the written register; this
may also remind one of the stream-of-consciousness manner (i.e. 'I did this . . . and then I did that . . . and next I . . . and my
brother did this . . . etc.)
Danish interference in the language of the compositions
In the field of lexical-semantic interference, examples of a direct borrowing of Danish words in the Serbo-Croatian language are
quite rare:
(Q.No.808)Imamo veliki reol.
(Q.No.837)Ja idem u dansku skolu*, idem u 9. razred
Hedegårdsskolen.

< previous page page_163 next page >


< previous page page_164 next page >
Page 164
Danish influence is reflected in the use of Danish syntagmas whose parts may be translated and taken over:
(Q.No.802)Sjedimo dobro i cisti * su stolovi . . . (under the
influence of 'Vi sidder godt . . .')

Most of the total of examples of interference are those pertaining to Danish grammatical interference in Serbo-Croatian.
The most common examples of morphological interference are related to misused (displaced) case endings or a reduction/loss of
them:
(Q.No.842)Kad dodem u razredu onda su puno od moji drugari i
drugarice napolje da igraju fudbal (under the
influence of 'Når jeg kommer til klasse lokalet er
mange af mine venner nede for at spille fodbold');
(Q.No.848)Ima cetiri decije* sobe na drugi sprat (probably
under the influence of 'på anden sal');

however, it is hard to tell whether the above (Q.No.848) is due exclusively to Danish interference (i.e. no declension of the
substantive), knowing that a construction such as this can be heard in some (eastern) Serbo-Croatian dialects.
Below are two examples of recurring syntactic interference, showing Danish word-order in Serbo-Croatian sentences:
(Q.No.816)Stara zgrada je preko sto godina stara, a nova zgrada
je 16-17 godina stara (under the influence of '. . . det
nye hus er 16-17 år gammelt.')

The repetition of the subject zgrada in two or more successive independent/ dependent clauses is redundant in Serbo-Croatian.
(Q.No.815)Onda sve to je doslo* u novinama (under the
influence of '. . . så kom det hele i avisen').

The pattern of possessive adjectives min, mit + noun(s), which is so common in Danish, has become adapted in Serbo-Croatian,
which does not normally favour it:
(Q.No.828)Tu zive* troje, moja majika i moj otac i ja (under the
influence of '. . . min mor, og min far og jeg');
(Q.No.846)Svakog dana idem u skolu* gde provodim najvise*
moga vremena.

At the morpho-syntactic level, an erroneous use of Serbo-Croatian prepositions is recurrent under the influence of Danish ones:
(Q.No.813)I nekada idemo na velike muzeje i tamo ima svasta*
(cf. 'Nogle gange tager vi på dit store museum . . .')

< previous page page_164 next page >


< previous page page_165 next page >
Page 165
Interference at the level of orthography is uncommon:
(Q.No.815). . . u boravku ima bilijar, tennis, stolni tennis,
nogomet, rukomet;
(Q.No.813). . . ko prvi slupa bure bude kralj ili kraljica teo dan
(Danish t is pronounced as /ts/; Serbo-Croatian
spelling of the word is ceo).

Having compared our respondents' language skill as exhibited in the multiple choice test with their written productions
(compositions), we have reached the conclusion that their passive grasp of the language (i.e. at the level of recognising correct
answers) is better than their active application of the internalised standard.
Factors that may stimulate L1 acquisition, maintenance and development
We have studied some factors that may motivate Yugoslav migrant children to learn and maintain their mother tongue.
The intention of migrant parents to return to their homeland and reintegrate in the society of origin, entailing their children's
return and, possibly, insertion in that community's school system may be an important factor influencing motivation for mother
tongue learning. The results of our research have confirmed again how difficult it is to resolve the migrant dilemma of 'whether
to stay on or to return'. It may not always depend on subjective aspirations, but is often resolved as a result of objective
circumstances or a contingency.
Parent interests or aspirations need not coincide with those of their children. Compromises and shortcomings are not uncommon
in a migrant worker's family, which may even split up (the old generation returning to the country of origin, the new one
embracing the adoptive community and its values). The planned strategy of life choices may be reflected in the language sphere.
Twenty of our respondents believe that their parents intend to return to the homeland, 6 do not think so, and 24 do not know. As
for themselves, 7 intend to continue their education in Yugoslavia, 23 do not, while as many as 20 do not know where they will
complete their education.
An important factor stimulating mother tongue learning motivation may also be migrant children's need to be identified as
members of their own ethnic group, their people, the country that they originate from, speaking 'their own' language. Forty-four
pupils (among them 18 who have spent 10-12 years in Denmark and 12 who have resided there 13-15 years) think of themselves
as Yugoslavs, one does not, four do not know whether they can call themselves that (one gave no reply to the question on their
national affiliation). Our premise to be tested by means of this ongoing research is that the respondents with an unstable national
identity may be less motivated

< previous page page_165 next page >


< previous page page_166 next page >
Page 166
to learn and maintain their mother tongue than those with a stronger national affiliation.
The factor that we would like to mention as last, but not least, important is migrant children's plans concerning their future life
in either of the two countries. Fifteen respondents want to live in Yugoslavia in the future (11 'because it is the homeland', four
'like it, love it' there), four want to remain in Denmark (three 'have adapted themselves so well', one 'for economic reasons'), two
want to live in a third country, as many as 29 are undecided.
Conclusions
The results of our survey conducted in Copenhagen in 1986 should be treated as preliminary; we shall be able to reach
conclusions after computer data processing has been carried out. However, on the basis of the frequencies of answers that have
been counted so far and of our examination of the respondents' compositions, the following statements can be made:
The sample of our respondents, 50 Yugoslav pupils in Denmark, is not significantly unrepresentative, although it does not
include non-attendants at supplementary mother tongue education. A vast majority of Yugoslav second generation migrants in
Denmark, as many as 43 (86%) are children who were born there. They have mastered the Danish language, attend regular
school, and are very good pupils. They have been accepted by their Danish peers. Almost half of them (48%) are members of a
school organisation, association, society or club engaging in sports, cultural, technical and sundry activities.
Our respondents have learnt most of their mother tongue from their parents, and they have been using it mostly with them, too.
In interaction with their siblings, 22 children prefer their mother tongue, but another 22 prefer Danish in the same circumstances.
As many as 38 pupils (76%) do not use their LI at play, but they speak Danish. The communication range of the mother tongue
is evidently narrow: in the host country it is generally used with migrant parents, relatives and compatriots and at mother tongue
classes. On the other hand, the majority language is used regularly, gradually replacing the mother tongue in all spheres of its
use.
In writing, recognition of mother tongue norms (i.e. in the multiple choice test) prevails over their active application (i.e. in
compositions). A comparison can be made with foreign language learning: when conversations are insufficient and reading is
meagre, a passive knowledge of the language cannot be readily brought to life.

Notes
1. Karmen Brcic * is responsible for the manual statistical processing of data that were collected for the purposes of the project
'The Language of Yugoslav children living in European

< previous page page_166 next page >


< previous page page_167 next page >
Page 167
countries', Andrina Pavlinic-Wolf * is the project leader and author of this article, and Nadezda Jeftic* has reported on
Danish-Serbo-Croatian interferences in the compositions of our Yugoslav respondents in Copenhagen.
2. During her stay in Copenhagen in the summer of 1986, A. Pavlinic-Wolf obtained this and many other impressions from
interviews she had with several Yugoslav citizens, who have been living and working in Denmark for quite some time.
3. The structure of the rest of the paper is similar to Pavlinic-Wolf et al. (1987).
4. The results of the survey are preliminary and should not be quoted without a prior consultation with the authors.
5. In view of the respondents' number (N= 50), it is easy to calculate percentages: by multiplying a given number of respondents
by two. Therefore, they are not to be given in the text unless deemed necessary.
6. For the sake of comparison: two control groups of Yugoslav pupils in Zagreb (two eighth forms, a total of 56 children, aged
14) were very successful at the levels of syntax and semantics (1.1% and 0.4% incorrect answers, respectively), while they made
most mistakes at the level of orthography (17.1%), and next at the levels of morphology (7.9%) and phonology (5%).
7. For the purpose of our study we decided on having a formal, and not a content analysis of the compositions, although they
might serve as a valuable source of information on Yugoslav children in Denmark to psychologists, sociologists,
psycholinguists, etc.
8. The girl (11) who wrote the composition was born in Copenhagen and had been living there all the time. She described an
interesting encounter with a hare that had been trying to run across a highway near Belgrade when she was taken for a night
drive in a car with her father and brother. When asked by the author of this paper, who addressed her very softly and tactfully,
why she had decided to write about the hare, she burst into tears and left the classroom without an explanation.

References
Document MMG-3(87)17 (1987) Migrants in Western Europe: Present Situation and Future Prospects. Strasbourg: Council of
Europe.
Document (1983) Informacija o nastavi maternjeg jezika i nacionalne kulture za jugoslovensku decu u inostranstvu. Belgrade:
Savezni biro za poslove zaposljavanja*.
Modersmåls undervisningen i Hovedstadsområdet (1985-87). København: Kontoret for fremmedsprogede elever.
Pavlinic-Wolf, A., Anic*, J. and Ivezic*, Z. (1987) Jezik jugoslavenske djece koja zive* u Svedskoj*. Child Language in
Diaspora (Serbo-Croatian in West European CountriesPapers from a Symposium). Slavica Lundensia 11, 159-173. Lund:
Slaviska Institutionen vid Lunds Universitet.
Schierup, C.-U. (1984) Do they dance to keep up tradition? Analysis of a social situation among Yugoslav immigrants in
Scandinavia. Research Report No.81. Umeå: Department of Sociology, University of Umeå.
(1985) Why are vampires still alive? Wallachian immigrants in Scandinavia. Ethnos. Xeroxed manuscript, Department of
Sociology, University of Umeå.
(1986) Options of unemployed immigrants in the 1980sa comparative study of unemployment among Turkish and Yugoslav
immigrants in Sweden and Denmark. Migracijske teme 3-4. Zagreb: Centar za istrazivanje* migracija i narodnosti, 77-86.
Schierup, C.-U. and Ålund, A. (1987) Will They Still Be Dancing? Integration and Ethnic Transformation among Yugoslav
Immigrants in Scandinavia. Stockholm: Almqvist & Wiksell.

< previous page page_167 next page >


< previous page page_ii next page >
< previous page page_169 next page >
Page 169

Bilingual Education in Yugoslavia:


Some Experiences in the Field of Education for National Minorities/Nationalities in Yugoslavia
Sonja Novak-Lukanovic *
Institute for Ethnic Studies,
Ljubljana 61000, Erjavceva* 26., Yugoslavia.
Abstract. Yugoslavia is a multinational, culturally pluralistic and multilingual society. The multilingualism is reflected not
only in the use of the languages of the national minorities/nationalities but also in the diversity of languages of the nations
of Yugoslavia. The equality of these languages is also reflected in the system of education, through which it is possible
for the members of various nations and national minorities/ nationalities to be educated in their mother tongue. In
individual republics and provinces, different forms of education have been developed in accordance with tradition and the
specific needs of life, in which the function of the individual languages is variously defined. In the nationally mixed
regions, depending on whether the minority language is mother tongue (L1) or second language (L2) in the curricula, the
following typology is used: Type A, minority language (L2) is the language of instruction; Type B, bilingual education:
Type C, minority language (L1) is the subject of instruction; and Type D, minority language (L2) is the subject of
instruction. The aim of education in the nationally mixed regions of Yugoslavia is, among other things, to develop the
motivation for equal use of the languages, so the language of national minority/nationality also becomes an appropriate
instrument for communication in public and social life and is not restricted to usage in private life. The system of
education represents only the first institutional level which leads to the development of bilingualism.

Introduction
Six nations live in Yugoslavia (Montenegrins, Croats, Macedonians, Muslims, Slovenes, Serbs), together with ten
nationalities/national minorities (Albanians, Bulgarians, Czechs, Italians, Hungarians, Romanians, Ruthenians, Slovaks, Turks
and Ukrainians) and two ethnic groupsGypsies and Vlachs. In addition to these, members of various other nationalities live
scattered throughout the entire territory of Yugoslavia, though not in large numbers.

< previous page page_169 next page >


< previous page page_170 next page >
Page 170
The 1981 census gave a total population for Yugoslavia of 22,427,585, of which 2,501,369 or 11.2% of the entire population are
members of various nationalities/national minorities, while the two ethnic groups represent 0.8%. These data demonstrate the
multinational, culturally pluralistic and multilingual nature of Yugoslav society. Multilingualism in Yugoslavia is reflected not
only in the use of the languages of the national/nationalities minorities but also in the diversity of the languages of the nations of
Yugoslavia.
The 'language policies', if such a term can be used to describe the series of constitutional and legal acts and other activities in
Yugoslavia, are based on the equality of all the languages of the nations and national minorities/ nationalities on all levels of
social life. The equality of these languages is also reflected in the system of education, through which it is possible for the
members of various nations and national minorities/nationalities to be educated in their mother tongue. Education in the mother
tongue is the practical effect of the Yugoslav model of multilingualism, which excludes any implication that one language may
be inferior or superior to another (Mikes *, 1984). Cultural pluralism, and Yugoslavia is a culturally pluralistic society, does not
permit the domination of one ethnic group by another, or any other form of domination of one culture over another (Klinar,
1986). And it is precisely for this reason that the education of minority members/ members of nationalities in their mother
tongue has an additional 'minority component', in that it helps them to find their identity and to educate them with the aim of
improving their socioeconomic position. Although any kind of education is political and socially conditioned, education relating
to ethnic and national minorities/nationalities is more than any other education deeply embodied in its social, political and
historical context (Cziko & Troike, 1984). Despite the uniform foundations of 'language policies' in the field of education for the
members of the national minorities/nationalities of Yugoslavia, different forms of education have been developed in the
individual republics and provinces in accordance with tradition and the specific needs of life, in which the function of the
individual is differently defined. Thus the typology of education for nationalities/national minorities does not originate only from
different theoretical or methodological approaches but also expresses the different specific historical, cultural and demographic
circumstances in the territory of Yugoslavia (Novak-Lukanovic*, 1986).

The Typology of Education for Members of National Minorities/Nationalities in Yugoslavia


The typology of education for the members of national minorities/nationalities can be defined on the basis of the role played by
the languages (the language of the minority or that of the majority, respectively) in the educational programmes and on the basis
of the objectives/aims set in the

< previous page page_170 next page >


< previous page page_171 next page >
Page 171
educational programmes within the individual republics and provinces. Bilingual education in Yugoslavia is always connected
with education in the nationally mixed regions. How can we define this term? The nationally mixed region is a legal fiction: an
area inhabited by members of a nation representing a majority and members of a national minority/nationality representing a
minority. The constitution and the statutes of the communes define these territories as being 'nationally mixed regions'. The
forms of education are various, but according to the role of the minority language being mother tongue (L1) or second language
(L2) in the curricula the following typology can be presented:
Type A:Minority language = mother tongue (L1) = language of instruction;
Type B: bilingual education;
Type C: minority language = mother tongue (L1) = subject of instruction;
Type D: minority language = second language (L2) = subject of instruction.
Type A: minority language = mother tongue (L1) = language of instruction
This type includes elementary schools in which the language of instruction is that of the national minority/nationality. The
acquisition of elementary education and literacy training in the mother tongueespecially for those who belong to national
minorities/nationalitiesare extremely important factors in promoting cultural life (Mike *, 1984). There have been many
declarations on the importance of education in the mother tongue. The following is that given by UNESCO:
It is axiomatic that the best medium for teaching a child is his mother tongue. Psychologically, it is the system of
meaningful signs that in his mind works automatically for expressions and understanding. Sociologically, it is a means of
identification among the members of the community to which he belongs. Educationally, he learns more quickly through
it than through an unfamiliar linguistic medium (UNESCO, 1953).
This type is the dominant pattern in many republics and provinces of Yugoslavia. In the republic of Slovenia, in the nationally
mixed region of Obala, Italiansas members of the national minority nationalityattend elementary schools in which the language
of instruction is Italian. In the republic of Croatia, Italians, Hungarians and Czechs can attend elementary schools in which the
language of instruction is their mother tongue (Italian, Hungarian, Czech). The same type of schools, in which the mother tongue
is the language of instruction, can also be attended by Albanians in the republic of Montenegro, Albanians and Turks in the
republic of Macedonia,

< previous page page_171 next page >


< previous page page_172 next page >
Page 172
and Albanians and Bulgarians in the republic of Serbia.
In the province of Vojvodina, the Serbo-Croat, Hungarian, Slovak, Romanian and Ruthenian languages have equal rights and
are all used as media of instruction. Also due to the variegated national make-up of the population in the province of Kosovo,
the language of instruction is Albanian, Serbo-Croat or Turkish. In this type of education, in which the language of instruction is
one of the languages of the national minorities/nationalities, the pupils are obliged to learn a second language (L2), the language
of the majoritythat is one of the languages of the nations of YugoslaviaSlovene, Serbo-Croat or Croat-Serbian or Macedonian.
Type B: bilingual education
The elementary school in the region of Prekmurje (north-east of Slovenia) is an example of a bilingual school in which Slovenes
and Hungarians attend the same school, without regard to their national affiliation or the wishes of the individual. The
importance of pre-school education must be stressed. Attendance at pre-school classes is compulsory for all children for two
years preceding entry to primary school (in other parts of Slovenia this is one year), where they are systematically introduced to
the second language (L2)Hungarian or Slovene and simultaneously develop their expressive capacity in their mother tongue.
This pre-school knowledge of the second language (L2) provides the basis for bilingual lessons when they enter primary school,
where both languages are school subjects and used as teaching languages. In the first stage (grades 1-5), the teacher explains
new learning material in each subject, first in Slovene and then also in Hungarian. Various forms of fixing, deepening and
testing proceed on the principle of internal differentiation and individualisation. In the higher grades (5-8), lessons in the natural
sciences and mathematics are mainly in Slovene (80%) as are lessons in history and geography. Only teaching units which
contain Hungarian history and geography are given in Hungarian, with additional Slovene terminology. So at the lower level,
bilingualism is complete, and at the higher level only partial. Such bilingual education (the educational process taking place in
two languages) is not common practice in other republics and provinces. However, there was marked interest in this type of
education in the province of Vojvodina; although the experiment produced satisfactory results, this type of education was not
introduced (Mikes *, 1984).
Type C: minority language = mother tongue (L1) = subject of instruction
In some parts of Yugoslavia, it is impossible to organise classes in the mother tongue or bilingual classes for the members of
national minorities/nationalities, so these pupils attend lessons in their mother tongue on an optional basis. The fact that Italians,
Ukrainians and Czechs live widely scattered, dispersed throughout the territory of Bosnia and Herzegovina,

< previous page page_172 next page >


< previous page page_173 next page >
Page 173
affects language development and so also the organisation of education. These pupils attend only hours of mother tongue
instruction, for which the term 'language cultivation' is used. Instruction in the mother tongue of the same type and for the same
reason as in Bosnia and Herzegovina is organised for the members of Hungarian, Czech, Slovak, Ruthenian and Ukrainian
national minorities/nationalities within Croatian or Serbian language schools in some communities of the republic Croatia.
Type D: minority language =second language (L2) = subject of instruction
In all the types mentioned in the paper (Types A, B, C) the members of the national minority/nationality learn, in addition to
their mother tongue (L1), the language of the majority as the second language (L2). But in nationally mixed regions in
Yugoslavia (Type D), the members of the majority who attend schools in their mother tongue or bilingual schools learn the
language of the national minority/nationality as a second language (L2). The learning of the second language (L2) is variously
organised in individual regions of Yugoslavia. Sometimes it is compulsory, sometimes it is an optional subject.

The Aims of Education in Nationally Mixed Regions


The intention behind education in the nationally mixed regions of Yugoslavia is, among other things, to develop motivation for
equal use of the languages, so the language of the national minority/nationality becomes an appropriate instrument for
communication in public and social life and is not restricted only to usage in private life. A characteristic of the realisation of
the teaching aims is that educational programmes specifically include aspects of the history and culture of the nations and
national minorities/nationalities with whom the pupils live. The members of the majority get to know the culture and history of
the minority, which can essentially assist in reducing psychological and other barriers which can occur, either as a result of
historical friction or having their source in contemporary socioeconomic, politico-psychological and other conditions which
prevail in particular surroundings (Necak-Lük *, 1986).
It is difficult to answer the question as to whether the educational process in Yugoslavia trains individualsmembers of the
national minority/ nationality as well as members of the majority populationsuch that they acquire in their schooling years the
capacity for bilingual communication and equal mastery of both languages, since the success of such a planned educational
process is influenced on the one hand by appropriate pedagogic methods and on the other by the attitude of society as a whole
(Cummins, 1986). All this influences the attitude of the individual towards bilingualism and his learning and knowledge of the
second language (L2). Successful educational methods, positive attitudes of society reflected in numerous

< previous page page_173 next page >


< previous page page_174 next page >
Page 174
constitutional-legal acts and laws and a positive attitude of the individual towards the second language (L2) influences the
development of functional bilingualism in a specific surrounding, which means that both the languagesthe language of the nation
and of the national minority/nationalityare equally used outside the school, in the home as well as in all spheres of social life
(Fishman, 1977).
How the institutionalised situation of the language in specific social surroundings influences the degree of knowledge or
function of individual languages is shown by an example from the republic of Croatia, where in certain regions settled by
Italians, who are not defined by statute as a national minority (e.g. Pula), the Italian language as the mother tongue (L1) or as the
second language (L2) is taught optionally. In addition, the individual's choice as to whether or not to learn Italian, to accept or
reject the status of bilingualism, is 'free' (Milani-Kruljac, 1984). In these circumstances, when the constitutional-legal situation
should guarantee the development of the nationality, individual bilingualism becomes a subjective choice and the Italian
language is in a subordinate position and is excluded from the social, political and administrative life of the region.
For the successful development of bilingualism as a social phenomenon, bilingual education cannot and must not stop at primary
school level. It is also important how the pupil who has finished primary school of Types A, B, C or D continues schooling or,
should we say, how he enters the work situation. The position of the language of the nationality as mother tongue or second
language (L1 or L2) plays an important role in further education and entering the work situation. It is entirely unrealistic and
illusory to expect nationally mixed regions to ensure all types and degrees of secondary and college education as this has been
presented for primary schools. Pupils wishing to follow different professions are directed into schools ouside their nationally
mixed regions, which is confirmed by data from the Republic of Slovenia (Novak-Lukanovic *, 1986). The role of the national
language as the mother tongue or second tongue is thus changed. In the Republic of Slovenia, despite legal possibilities for the
pupils to continue their post-primary education in the language of the nationality, also outside the nationally mixed regions, this
possibility becomes a subjective choice which greatly depends on the motivation of pupils towards the learning of the language
of the nationality. This is influenced not only by the national affiliation of the individual, but also by attitudes within the micro
and macro surroundings.
Some studies on motivation and attitudes towards learning exist (Strukelj*, 1979). This contribution will mention only a few
such motivations which, at least in our opinion, influence learning and knowledge of the language of the nationality and thus the
development of the equal use of two languages, which also represents one of the aims of education in nationally mixed regions
in Yugoslavia. These motivations include:
motivations which derive from the national affiliation of the individual;

< previous page page_174 next page >


< previous page page_175 next page >
Page 175
the socioeconomic motivation of the individual; and
the individual's motivation for the acquisition of knowledge.

The first motivation mentioned springs from the national identity of the individual and is present in members of nationalities, to
whom the learning of the language of the nationality means the learning of the mother tongue (L1). The socioeconomic
motivation means that knowledge of the language of the nationality is very important in providing communication between
speakers of different nationalities. Knowledge of the language gives the basis for the creation of co-existence and represents a
very important element, one could say even precondition for the individual's participation in life it makes it possible for him/her
to enter work and thus gives economic security and provides the possibility of progress in his/her profession. This
socioeconomic motivation is usually related to certain professionsteachers, employees who work with the public, traders. The
socioeconomic motivation of the individual is also influenced in our opinion, in a negative manner, by the principle of territorial
limitation of bilingualism.
Similarly with the language of the nationalitywhich linguistic group it belongs to, whether it has the status of an international
language, what is its role in the world and what relations it has to the state (country) of the mother nation in political, cultural
and economic fieldsall these are aspects which can affect, in a negative or positive sense, the socioeconomic motivation of the
individual for learning the language of the nationality. The motivation for acquisition of knowledge can be defined as a purely
linguistic motivation, whose aim is not learning the language of the nationality as the mother tongue or second language in order
to communicate with people in the common environment, or to get to know their culture and historythus creating co-existence in
a specific surrounding. The individual learns the second language primarily as an investment in his own knowledge, the
acquisition of 'mankind's intellectual capital' (Grenier & Vaillancourt, 1983). Of course, in the majority of cases, this motivation
touches on the socioeconomic motivation of the individual.

Conclusion
Bearing in mind various definitions of when an individual is bilingual (Ovando & Collier, 1985), we may claim that in
nationally mixed regions, members of the nation and members of the nationality are both bilingual, although different levels of
bilingualism can be distinguished. The results of surveys in nationally mixed regions in Slovenia indicate that in contacts with
the public, in offices and places of work, the language of the majority is usedSlovene. This raises the question of why, despite
objective possibilities (educational system) and positive factors (attitude of society), which encourage the use of the languages of
both the majority and the minority, in certain formal contacts the use of the majoritySlovenelanguage predominates. There may
be many reasons and answers, and we

< previous page page_175 next page >


< previous page page_176 next page >
Page 176
shall mention only a few which derive from surveys and studies. One, at least in our opinion, very indicative pointer (Dular,
1986; Necak-Lük *, 1983) is that, in the development of society, the majority nationality has also developed a specific self-
management terminology, which does not exist in the language of the nationality and which is also difficult to translate. This is
why in some situationsabove all at conferences, in places of work evenmembers of a nationality themselves decide to use the
language of the majority, Slovene. Very often entirely psychological factors intervene (Miklic*, 1984), when the speaker
(meaning a member of the majority, a Slovene) is actively aware of the inadequacy of his linguistic knowledge and his inability
to express suitably in a new code (L2) the content of his message. As he is aware that the co-speaker also understands his
mother tongue, he uses Slovene, although he has been trained in his years of schooling to communicate in other languages (L2).
In conclusion, I should again stress that the aims of curricula which are applied in nationally mixed regions in Yugoslavia are
realised by organisation, the content and pedagogic method, but they are also influenced by numerous social factors.

References
Cuminins, J. (1986) Empowering minority students: a framework for intervention. Harvard Educational Review, 1, 18-36.
Cziko, G.A. and Troike, R.C. (1984) Contexts of bilingual education: international perspectives and issues. ILA Review, 1, 7-34.
Dular J. (1986) Jezikovni polozaj* v Dobrovniku. In Madzari* in Slovencisodelovanje in sozitje* v obmejnem obmozju* ob
jugoslovansko-madzarsk*i meji. Ljubljana: INV.
Fishman, J. (1976) Bilingual Education: An International Sociological Perspective. Rowley, Mass: Newbury House.
Grenier G. and Vaillancourt, F. (1983) An economic perspective on learning second language. Journal of Multilingual and
Multicultural Development, 6, 471-83.
Klinar, P. (1986) Socioloski* vidiki multikulturaliszma. Migracijske teme, 2, 73-84.
Mikes*, M. (1984) Instruction in mother tongues in Yugoslavia. Prospects, 1, 121-31.
Miklic, T. (1984) Komunikacija v drugem jeziku in jezikoslovje: ali jezikslovje lahko prispeva k razvijanju in utrjevanju
dvojezicnosti. In Dvojezicnost-individualne* in druzbene razseznosti, Ljubljana: DUJS, 245-50.
Milani-Kruljac, N. (1984) Bilingualismo e statuti comunali: la siituazione Istro-Quarnerina. In Dvjezicnost-indviduale in
druzbene razseznosti, Ljubljana: DUJS, 47-53.
Necak-Lük, A. (1983) Druzbene razseznosti dvojezicnosti na arodnostne mesanem obmocju* Prekmurja. Doctoral dissertation,
Filozofska fakulteta, Ljubljana.
(1986) Education in multicultural societies and its societal implications. Razprave in gradivo, 18, 309-13.
Novak-Lukanovic*, S. (1986) Some Yugoslav experiences in asserting equality of the nations and nationalities in the field of
education. Razprave in gradivo, 18, 32-89.
(1987) Dvojezicna* osnovna sola*: kam in kako potem?. Ljubljana: Ekspertiza INV.
Ovando, C. J. and Collier, V. P. (1985) Bilingual and ESL Classrooms, New York: McGrawHill.
Strukelj*, I. (1979) Motivacija in stalisca* do ucenja* in rabe L1 in L2. Ljubljana: Institut za sociologijo. Univerze Eduardo
Kardelje.
UNESCO, (1953) The Use of Vernacular Languages in Education. Monographs on Fundamental Education VII. Paris.

< previous page page_176 next page >


< previous page page_177 next page >
Page 177

Language Policy and Language Planning after the Establishment of the Home Rule in Greenland
Aquigssiaq Møller
Kultur-og undervisningsdirektoratet, Grønlands Hjemmestyre,
3900 Nuuk, Grønland.
Abstract. When Home Rule was introduced in Greenland in 1979 the question about the languages was dealt with in the
Home Rule Act and in the Statutory Instrument of the Parliament, No 6 of 16 October, 1979, regarding the primary
school, in which it is stated in § 1.2 and 1.3 that the educational language is Greenlandic, and that Danish may be the
educational language if exceptional circumstances make it necessary.
For pupils with the Greenlandic language as their mother tongue, the subject of Danish may be introduced in the second
grade, and has to be introduced by the fourth. For other pupils special training in Greenlandic as a foreign language is
given.
It must be stated that the efforts of strengthening the Greenlandic language and making it the principal language, within
the administrative field as well, have been fruitful. But it has been a slower process than expected. This is due to many
things which I shall deal with in this paper.
The Home Rule Act and the Statutory Instrument about the primary school very clearly state which language is the main
language. The situation is the same for the media policy, where the aim is a distribution between the languages of 80%
Greenlandic and 20% Danish.
Recent Greenlandic school history and the debate about education and learning is very much centred on language
problems. There is a tendency to discuss the acquisition of language as a final purpose in itself instead of as a means to an
end.
The aim is bilingualism in Greenland, oras stated by the Home Rule Actthat Danish has to be mastered well and
thoroughly.
The situation of the language policy and planning after seven years of Home Rule is illustrated by information about
debates in the Parliament, in parents' assemblies and in teachers' assemblies. Summary by Gerd Gabrielsen
The Home Rule Act of 1979 instituted a bilingual language policy by stating that Greenlandic is the main language of
Greenlandic society, and that a high level of proficiency in Danish will be generally required. This policy is intended to cover all
areas in Greenlandic society. Its purpose is to preserve and further develop Greenlandic culture.

< previous page page_177 next page >


< previous page page_178 next page >
Page 178
In administration, bilingualism throughout is aimed at, but in certain areas, e.g. legislation, for practical reasons texts are still
written in Danish and then translated into Greenlandic. For the media, TV and radio, 80% Greenlandic-language and 20%
Danish-language coverage is the aim, while written media which receive support from the Home Rule administration are
bilingual throughout. The role of the media in language development and language change can hardly be over-estimated. The
information made available through the media is a means of conceptualisation as well as a prerequisite for the process of
democracy.
In education the aim is for Greenlandic to be the main language throughout. Flexible practices, however, are made necessary by
practical considerations such as a lack of Greenlandic-language or bilingual teachers, materials, etc. For schools in general there
are four main areas of development:
1. Provision of bilingual teachers;
2. Development of curricular guidelines, teachers' guidelines and materials, i.e. development of teaching at the level of aims,
methods and educational experience in all school subjects;
3. Development of parents' and pupils' motivation for education in/through school;
4. Development of a general awareness in society of the importance of language along the dimensions of
language/communication, language/society and language/political action in a bilingual society.

As to languages, the development of Greenlandic as a first language is of primary importance, both as to form/conceptual
coverage and at a functional level. Danish is important as the first foreign language. There is at the moment an ongoing process
of change, the outcome of which it is too early to estimate linguistically and qualitatively. The relative balance between the two
languages as measured in the number of school hours given is being changed: the number of hours spent in Danish is being
decreased, while the number of hours spent in Greenlandic is being increased.
A general improvement in the teaching of Danish is currently being attempted:
1. New curricular guidelines have been developed, while materials are under development, taking into account the present, and
not least the future, use of Danish as a first foreign language and the particular use made of the Danish language in Greenland.
(Material will be completed within two or three years for grades 1 to 9.)
2. A radical change in or the abolition of present Danish-based exams.
3. Intensification of teaching and in-service provision for Danish teachers.

At the present time, after eight years of Home Rule administration, the debate on languages in school is polarised, with a clear
trend towards favouring the development of Greenlandic. In basic general schooling it is

< previous page page_178 next page >


< previous page page_179 next page >
Page 179
possible for parents to enter their child in either a Greenlandic-based or a Danish-based stream, depending on the home
language of the child. In Greenlandic-based teaching Greenlandic is the rule, with the proviso that Danish teachers may be used
if no Greenlandic or bilingual teachers are available. Danish is taught as a foreign language from the fourth year. In Danish-
based teaching, foreign language classes in Greenlandic are given for two hours a week.
Recently, however, arguments have been put forward for a third position, reflecting the fact that there are in Greenland basically
three sociolinguistic groups: Greenlandic, Danish and mixed. This is interesting and may prove influential in future
developments in schools.
Parents protest against having to choose either a Danish-based or a Greenlandic-based education for children who in their home
background have established bilingual and bicultural identities. It is argued that neither choice is satisfactory and that, with
purposive planning, these children might develop full bilingualism in school. A change in this direction will require a general
change of attitude, and changes in planning and teaching.
The question of the place of English is frequently raised. At present, English is not compulsory. It is a clear wish, however, on
the part of both politicians and parents, that English should be introduced as a compulsory subject in the sixth or seventh grades,
as a means of communication with the world at large and with the Inuit populations of Canada and Alaska.
A review of language conditions in Greenland shows that with all the changes the one constant element is the land, the country
and the conditions for living it offers now and not least in a long-range perspective.
Bilingualism is a necessity in Greenland. The problems this raises are complex and will find no easy solution. Practical issues
will change with situation, time, political and pedagogic intentions. However, the attempt to develop Greenlandic as an effective
language in the fields of politics, administration and culture is based on unanimous consensus. The teaching of Danish will be
developed to suit the purposes and uses of that language in a Greenlandic context.
The finaland decisiveissue is the child. Will bilingualism for the child be a burdenpossibly destructiveor will it work as a
challenge to further a better understanding and better conditions in Greenlandic society? The East Greenlander Asineq says of
words:
The word is the greatest power human beings have. With words you can wound others or make them happyfor life.
If one is wounded by a weapon, and the wound heals, it can still be seen, but it no longer hurts as does the word which
has once been pronounced. Therefore the word is the greatest power of human beings.
We believe in Oqaatsip Kimia'the power of the word'. A wondrous power. The word is magic.

< previous page page_179 next page >


< previous page page_ii next page >
< previous page page_181 next page >
Page 181

Linguistic Minorities and Language EducationThe English Experience


Euan Reid
Department of English for Speakers of Other Languages,
University of London Institute of Education,
20 Bedford Way, London WCIH OAL, U.K.
Abstract. The paper begins by giving briefly some information about the main linguistic minorities in England, including
the kind of sociolinguistic interaction patterns that are common. It goes on to outline the three major stages of response by
the public education system to the presence of large numbers of speakers of other languages in English schools. The final
section gives a critical summary of the current situation and poses some key questions for discussion.

Linguistic Minorities in England


To understand the nature of the educational response to linguistic minorities in any school system, it is important to know
something about who the minority pupils are, where they (or their parents or grandparents) came from, and what their status in
the new country isfor example, in terms of citizenship.
The easiest way to do this for England, since there are no questions about language in the National Census, is to look at
information deriving from some surveys of languages in schools done over the last ten years or so. Figure 1 presents four sets of
data from different regions of England, and taken together these give some indication of the linguistic make-up of the school
populations for which policies in England have been designed. (For further details, see Linguistic Minorities Project, 1983,
1985; ILEA, 1985.) The circles in Figure 1 represent all the bilingual pupils in the schools of the different areas. Their numbers,
as compared with the whole school populations, are indicated beside each of the circles.
Bradford, a once-prosperous wool town in West Yorkshire, is fairly typical of many of the Northern and Midland conurbations
in terms of its linguistic

< previous page page_181 next page >


< previous page page_182 next page >
Page 182

Figure 1: *
Language in schools
(Sources: (a) Linguistic Minorities Project, 1985, p.328; (b) ILEA, 1985)

< previous page page_182 next page >


< previous page page_183 next page >
Page 183
make-up. The dominant languages are those from South Asia, in Bradford's case above all Urdu and Panjabi from Pakistan,
Kashmir and to a lesser extent from Indian Panjab. Speakers of these languages were recruited in the late 1950's and early 1960's
as cheap labour for the declining mills. Since they were mostly Commonwealth citizens, their families were able to join them, so
that by the mid-1960's the schools were taking in quite large numbers of 'non-English speakers'.
Peterborough is an East Midlands town in some respects quite similar to Bradford. Again Panjabi and Urdu are prominent
minority languages there, but so is Italianspoken now by the children and grandchildren of the original, mostly Southern Italian
migrants who came to Bedfordshire and Cambridgeshire, often to work in the very unattractive conditions of the brickworks in
the post-World-War-Two period. Linguistic minorities represent a much smaller proportion of the school population in
Peterborough compared with Bradfordabout 7% compared with about 18%.
In Haringey, in North London, the picture is rather different. First of all, the South Asian languages are much less prominent,
although also present: the major languages in this area are Greek and Turkishboth spoken mostly by people whose families came
from the island of Cyprus. Some have been in the U.K. for several generations already, although there was a sharp increase in
the mid-1970's, after the Turkish invasion of the island. However, in spite of the fierce antagonisms on Cyprus itself, there is a
high degree of at least economic interdependence between Greeks and Turks in North London. Together, they dominate some
sectors of the clothing industry, for example, with a high proportion of women used as home-workers.
Finally, in this brief selection, most prominent among bilingual pupils in Inner London as a whole are speakers of Bengali, or
rather of Sylheti, a variety of Bengali spoken in the North-Eastern corner of Bangladesh where nearly all the migrants come
from. Sylheti speakers have lived until recently mostly in the East End of London, but are now also in neighbouring areas. They,
too, are prominent in the clothing trade, and have on the whole much the worst working and housing conditions of all the
migrant groups in England.
The next most frequently reported language in Inner London is again Turkish, followed by a large range of other language
groupsmore than 150 in total, with the largest numbers being speakers of South Asian and South European languages.
As far as patterns of linguistic interaction are concerned, Table 1 comes from some quite small-scale surveys which the
Linguistic Minorities Project did in the early 1980's, with the intention of going beyond the large-scale language census and
survey data which are shown in Figure 1. Although it has to be interpreted with caution, given the nature of the school setting in
which the questions were asked, it shows rather clearly the kind of intergenerational differences that are beginning to appear in
some at least of the linguistic minorities in England.

< previous page page_183 next page >


< previous page page_184 next page >
Page 184
In the top half of Table 1, the first rather obvious point to notice is the shift towards English from one generation to the
nextmost dramatically shown in the differences between the child's language use with siblings as compared with parents. In the
bottom half of the table, another significant aspect is the non-reciprocal language use implied by the contrast between the top
and bottom of the table: for example, between parents and children. In other words there are clearly a good many conversations
in which mother or father uses the minority language to the child, who understands it, but replies in English.
Such complexities of interlanguage switching and mixing are confirmed if we look also at patterns of adult language use. Figure
2 includes two sets of data from the Linguistic Minorities Project's Adult Language Use Survey in London in 1981, first among
Italian households, then among Turkish speakers. In both cases, for high proportions of the samples it is not a simple matter of
choosing in a particular setting either English or the mother tongue, but of using both, in varying proportions (Reid, Morawska
& Smith, 1985).

Figure 2:
Household language use among bilingual adults in London
(Source: Reid, Morawska & Smith, 1985: 27, 51)

< previous page page_184 next page >


< previous page page_185 next page >
Page 185
Table 1 Linguistic interaction among some bilingual teenagers
Peterborough Bradford
% % % both % % % both
minority minority
N languageEnglishequally N languageEnglishequally
To my father I usually speak: 129 51 17 32 203 71 16 13
To my mother I usually speak: 132 62 13 25 209 78 11 11
To my brother(s) I usually speak: 112 16 50 34 200 19 60 21
To my sister(s) I usually speak: 107 17 49 34 189 25 54 21
To my grandfather(s) I usually speak: 83 74 16 10 127 87 6 6
To my grandmother(s) I usually speak: 95 78 13 9 130 78 6 16
My father usually speaks to me in: 128 62 16 22 203 76 14 9
My mother usually speaks to me in: 133 70 14 16 206 87 8 5
My grandfather(s) usually speak to me in: 83 82 14 4 131 87 8 5
My grandmother(s) usually speak to me in: 95 82 11 7 132 88 9 3
In school breaks, my friends and I usually 131 3 81 16 202 9 84 6
speak in:
Source: Linguistic Minorities Project (1985; p. 366).

< previous page page_185 next page >


< previous page page_186 next page >
Page 186
The Educational Response
Table 2, ESL in England 1962-87, represents one view of the last 25 years, divided into three main phases, roughly 1962-70,
1970-77 and 1977 to the present. Of course, the phases have in fact happened at different times in different parts of the highly
decentralised school system in England. My dates probably represent most accurately 'progressive practice', particularly in the
London area, rather than being generally true for the country as a whole.
The headings of the other columns for each phase taken together give a fairly comprehensive picture of the institutional
framework within which the language education of migrant children in England has developed during this period.
Phase 1: 1962-70
In the 1960's the initial response of local education authorities in most parts of the country to the arrival of immigrant children
was to set up separate language centres or 'induction centres'. In these, children who were
Table 2: ESL in England 1962-87
Phase Learners Institutions Materials Journals &
Associations
1962-70 'Immigrant 'Induction Structurally ATEPO 1
children' centres' based 'English for
or or immigrants'
'non-English 'language from foreign ATEPO 2
speakers' centres' language
[No MTT] teaching
or EFL models
1970-77 'Ethnic minority' 'Withdrawal Thematically 'Multiracial
or groups' based school'
'Linguistic 'mixed classes' NAME 1
minority' [MTT only] from primary New
pupils out of school] school 'topic' Approaches
models to
1977 'Bilingual 'Mainstream' Curriculum Multiracial
onwards learners' or content-based Education
or 'Regular NAME 2
'Black Asian classes' science, (National
(etc.) [some MTT history Antiracist
British' in school] (etc.) Movement in
Education)
MTT = Mother Tongue Teaching.

< previous page page_186 next page >


< previous page page_187 next page >
Page 187
often referred to as 'non-English-speakers' spent up to two years. They were separated from their English-speaking peers,
ostensibly so that they could be taught English to a level which would allow them to join classes in ordinary schools, but also, of
course, to satisfy majority parents that their children would not be 'held back' by the presence of large numbers of immigrant
children in the same classes.
In such centres there was usually also some teaching of subjects other than English, at least at the secondary stage, but the
curriculum was much narrower than in regular schools, and the teaching staff rarely included a full range of specialist teachers.
The teaching materials developed specifically for use in such contexts were derived essentially from foreign language teaching
models of the kind prevalent in the 1960'sthat is to say, they were language-structure-based, but were given settings appropriate
to the inner urban areas where most migrants lived on first arrival in the new country.
Teachers who specialised in this work in England formed themselves into an organisation called 'ATEPO', an acronym which at
first stood for the 'Association for the Teaching of English to Pupils from Overseas' ('ATEPO 1' in Table 2), but which was later
changed to the 'Association for the Education of Pupils from Overseas' ('ATEPO 2'). The name change represented a significant
broadening of the scope of the organisation, but still implied a focus entirely on 'them' and not yet on 'us'. ATEPO's journal was
known for its first four years as 'English for Immigrants', although even under that title its range of topics was much wider than
language.
All of this seemed at the time entirely a matter of common sense, and indeed in several parts of the country still seems so to
those responsible for administering the schools. However, at the end of 1986 the UK Commission for Racial Equality published
the result of a thorough investigation into a system of this kind, in Calderdale, Yorkshire (CRE, 1986). They presented a
formidable case that the system was discriminatory in effect (whatever the intentions behind it were), and in several quite serious
ways it was clearly in breach of both educational and race relations legislation. The view of most people in the field now is that
this 'Calderdale Report' will mark the belated end of Phase 1 throughout the country. Separate provision can no longer plausibly
be presented as equal provision.
Phase 2: 1970-77
By the early 1970's more attention was being given to meeting the needs of what were by then often referred to as 'ethnic
minority' or 'linguistic minority' pupils in the context of ordinary schools, at least at the primary stage of education. Sometimes
this was still in 'withdrawal classes': children were taken out from their ordinary class for an hour or more each day, or twice a
week, etc., and given special tuition in English. Meanwhile, their classmates were doing something else. The teaching materials
used in such classes overlapped considerably with those from Phase 1.

< previous page page_187 next page >


< previous page page_188 next page >
Page 188
However, the most significant move during Phase 2 was towards trying to teach children in 'mixed classes'i.e. linguistically
heterogeneous classes consisting of native English speakers, children whose parents had come from the Caribbean and who
spoke some kind of Creole-influenced variety of English, and speakers of other languages. In terms of learning materials, this
implied a shift away from teacher-centred, structurally-based foreign language-type materials, and towards child-centred
thematically-based materials, usable in 'mixed ability' situations. The origins of this approach to language teaching for linguistic
minority children lie much more in good mainstream primary school practice, using 'centre of interest', thematically-based
approaches to learning. During this phase 'ATEPO 2' changed into 'NAME 1'the National Association for Multiracial Education,
and its journal became 'Multiracial School'. This implied, of course, a shift of attention away from an exclusive pre-occupation
with 'them', and a fuller recognition that an appropriate response would also have to involve changes in the cultural and
linguistic education of 'us'the Anglo-Saxon (or 'Anglo-Celtic') majority.
In this phase, as in the previous one, almost no support was given to the learner's mother tongue, at least not in the mainstream
curriculum. At best, mother tongue teaching was given a kind of marginal status during the school day, with classes meeting
irregularly in lunch hours or after school, taught voluntarily by bilingual teachers, who were quite often not language-teaching
specialists themselves.
Phase 3: 1977 to the present
By the beginning of the third phase, which had begun in most areas by the mid- to late 1970's, a further significant shift was
beginning to take place. It was marked in part by changes in terminology: the expression 'bilingual learners' came to replace the
term 'second language learners' or 'linguistic minority pupils' from Phase 2. (Occasionally, although not generally, you will find
now terms like 'Black British' or even 'Black Asian British'.)
Even more importantly, perhaps, Phase 3 is marked by changes in institutional arrangements. Local authorities begin to close
even their secondary-level language centres, and to put more and more emphasis on what has become known as
'mainstreaming'the placing of all learners, even newcomers, into regular classes in all subjects of the school curriculum, from a
very early stage in their education. This is usually accompanied by the development of 'language support' roles for ESL teachers,
and, more recently, of 'collaborative learning' models. Separate language-learning syllabuses and materials have become
increasingly rare, and are replaced in effect by the curricula and materials of the standard curriculumin science, history,
technical drawing, geography, etc.

< previous page page_188 next page >


< previous page page_189 next page >
Page 189
Meanwhile, the journal Multiracial School has turned into New Approaches to Multiracial Education. That is to say, it is no
longer in principle just concerned with schools which are themselves ethnically mixed, but with all schools everywhere in the
country. 'NAME 1' in 1985 becomes 'NAME 2'the 'National Antiracist Movement in Education.'
Finally, in the last four or five years at least, 'mother tongues' or 'community languages' begin to be given some recognition, use
and even teaching in the mainstream schools themselves, and are no longer left entirely to the voluntary sector. However, this
practice is still not widespread, and is approved of by the Department of Education and Science only in the very early stages of
schooling, and again, rather grudgingly, in the secondary modern-languages curriculum.

Current Practice
At this point in the conference talk, extracts from video films which have been produced for use in teacher training courses were
shown. Details of these films and of where they can be obtained are given at the end of the paper.

Summing-up and Conclusions


The current situation can then be summarised under these three headings:
a) 'Separate' ESL classes and learning materials are becoming increasingly rare; they are being replaced by 'English Support'
for Bilingual Learners, provided in the context of mainstream classes at both primary and secondary school level; or, very
recently, by 'collaborative learning' or team teaching.
b) 'Mother Tongues' or Community Languages (i.e. languages of ethnic minority communities), are used occasionally in the
early years of education, essentially to help transition into fully English-medium schooling; their development in the middle
years of schooling is left largely to voluntary, out-of-school efforts; they may be re-introduced later as possible 'examination
subjects' in the last few years of compulsory education; they remain essentially marginal.
c) To complement the previous two elements Language Awareness courses are now frequently advocated, less frequently
actually delivered. They are intended for all pupils in schools, monolingual and bilingual alike, and sometimes linked to new
approaches to foreign language education. They occasionally incorporate material dealing directly with individual bilingualism,
societal multilingualism, and linguistic diversity in general.

< previous page page_189 next page >


< previous page page_190 next page >
Page 190
(a) + (b) + (c) represents approximately the Swann Report positionso-called after the name of a widely quoted enquiry into the
education of ethnic minority pupils (DES, 1985). It probably satisfies 'the liberal consensus' for the moment.
The same basic ideas underlie practice in both primary and secondary schools, and to some extent in both second language and
mother tongue teaching: earliest possible integration of language learners into the regular curriculum, emphasis on small-group
learning, emphasis on active role for the learners, and on collaboration between learners.
There are at least two obvious limitations in that position. The first and most striking feature, in a wider Northern European
perspective, is the almost total absence in England of any form of full bilingual education, in the sense of education through the
medium of more than one language. The MOTET Project illustrated in one of the video extracts shown at the conference has in
fact been so far the only moderately substantial, publicly-funded experimental programme in England involving teaching in any
language other than English: it lasted for approximately one school year, with about 50% of the teaching being in Panjabi, 50%
in English. One interesting feature of the presentation of the Project's results is the noticeably defensive tone adopted by the
researchers (MOTET, 1981).
Apart from MOTET, bilingual education has hardly even been discussed as a serious option, and will, after the 1987 British
General Election, probably be even less seriously considered. It is still the case that the overwhelming weight of resources,
above all of teachers, goes into teaching the 'second language', English.
The second very noticeable feature, as compared for example with the situation in the U.S.A., is the almost total absence of a
research basis for the practice described in this paper. It has to be said, however, that looked at with an appropriate degree of
scepticism, the influence of political ideology on the decision-makers in North America looks considerably more important than
the subtleties of the research findings. That would also be true of England: the ideology is perhaps even more apparent there,
because few feel it necessary even to pretend that policies have a research basis.
So far in England, then, there has been no substantial research effort in the field of bilingual education, and no large-scale
educational resources devoted to anything except ESL. If this remains the position for much longer, it seems likely that attempts
to maintain anything other than 'symbolic' bilingualism among settled, non-indigenous minority populations in England are
likely to have only marginal effect.
For any significant progress in this direction, at least two pretty unlikely developments would be necessary:
1. some kind of official or public status and tangible economic value for the languages other than English would have to be
signalled, to match the existing symbolic and 'solidarity' value which they have;

< previous page page_190 next page >


< previous page page_191 next page >
Page 191
2. some full, non-transitional, bilingual education programmes would have to be piloted and introduced in appropriate areas.
The Danish Minister of Education, in his introductory remarks to the Fifth Nordic Conference on Bilingualism, indicated rightly
that it would be a matter of great regret if the cultural resource represented by the presence in Northern European schools of
large numbers of speakers of non-indigenous languages was not effectively developed. England certainly has a long way to go
before it can be seen as meeting this challenge effectively, in either general social or specifically educational terms.
References
CRE (Commission for Racial Equality) (1986) Teaching English as a Second Language: Report of a Formal Investigation in
Calderdale Education Authority. London: CRE.
DES (Department of Education and Science) (1985) Education for All: Report of the Committee of Inquiry into the Education
of Children from Ethnic Minority Groups. London: HMSO.
ILEA (Inner London Education Authority) (1985) The 1985 Language Census. London: ILEA.
Linguistic Minorities Project (1983) Linguistic Minorities in England. London: University of London Institute of Education.
(1985) The Other Languages of England. London: Routledge & Kegan Paul.
MOTET (Mother Tongue & English Teaching Project) (1981) Summary of the Report. Bradford: The University.
Reid, E., Morawska, A. and Smith, G.P. (1985) Languages in London. London: University of London Institute of Education.
Video extracts shown were from two Inner London Education Authority Projects:
'SLIPSecond Language Learners in Primary Classrooms;
'SLIM' Second Language Learners in the Mainstream (of secondary schools).
Further details are available from the ILEA Centre for Learning Resources, 275 Kennington Lane, LONDON SE11 5QZ,
England.

< previous page page_191 next page >


< previous page page_ii next page >
< previous page page_193 next page >
Page 193

Latest Developments in Early Bilingual Education in France and Southern Europe.


Klaus-Erich Gerth
World Information Center on Bilingual Education (CMIEB),
2 rue de Logelbach,
75017 Paris, France.
Abstract. Language and culture teaching in primary schools in Spain, Italy and France has greatly evolved during the last
ten years.
As an example, we can note that in Spain (Catalonia) the autonomy process allowed the emergence of varied bilingual
education systems on a macro-social level with the intervention of the usual pedagogical means as well as audiovisual
media and new technologies.
In France, the evolution of some regional languages (Breton, Basque, Catalan, Occitan . . .), of languages and cultures of
migrant populations (Arab, Portugese . . .), but also of dominant foreign languages in the educational system brings about
new problems in centralised national education. That could lead to many changes in language policy outlook.
In Italy, an important step seems to have been taken with the passing of the bill on one compulsory foreign language in all
primary schools (age 7 or 8) from 1989 on.
The purpose of this paper is to present and analyse those changes in their own context. Furthermore, the paper deals with
the scientific and pedagogical implications of those early bilingual education developments.
Finally, some prospects are presented on the short-term possible evolution in this particular field of education in France
and Southern Europe.

Introduction: Three Major Trends


I cannot give here a detailed presentation of all these countries and the different situations in each part of Spain, Portugal or
Greece. So I choose Spainespecially CataloniaItaly and France, as we find in these countries the three major trends of bilingual
education in Western Europe. With Basque, Galician and Catalan, Spain stands for the first main tendency: re-appropriation of
their own language, culture and identity by national minorities. With a new law concerning the extension of the teaching of one
foreign language to primary level for all schools, Italy stands for the second

< previous page page_193 next page >


< previous page page_194 next page >
Page 194
fundamental trend: the evolution of monolingual majorities in their attitude to foreign languages and cultures. The case of France
is more complicated. If the problem of the integrationwhat form of integration?of immigrants' children in the quite centralised
national educational system has to be considered as a significant example of the third trend in the field of bilingual education in
Europe, France also has the privilege of being facedto a certain extentwith the other two main tendencies.
The last part of my paper will deal with some new research which could possibly lead to new perspectives in language planning
and policy in the near future.

The Historical Case of Catalonia


The civil war and the success of Franco had a damaging effect on the renewal of educational movements which were quite
important in Catalonia under the second Spanish Republic between 1931 and 1936. The Catalan language and culture were
suddenly considered subversive and as such totally banned from school and public life. During the first period of Franco's era, a
number of Catalonia's teachers maintained the tradition of their language in clandestine summer schools. At the end of the
1960's and in the 1970's these schools obtained an increasingly official status, so that the five or six teacher training centres of
Catalonia could insert their minority language and culture into nearly all their teacher training programmes. Two major facts
could explain this evolution:
1. As Charlotte Hoffmann (1987) wrote, 'To be a true Catalan has always meant that one must speak Catalan. And the
combination of language and perceived national character constitutes the ''uniqueness" of Catalonia'.
2. With 6.5 million inhabitants, its economic strength, its sociological composition and its geographical position, this region
obtained a considerable political independence from the central government and the Cortes with a statute of autonomy chosen
by 85% of the votes in 1979. In 1983 Catalonia's President Pujol, according to Article 32 of the new Statute of the Nation,
promulgated a law concerning the 'linguistic normalisation of Catalonia'. This law says, 'Catalan is from now on considered as
the natural instrument for the communication and expression of the people'. It is 'the symbol of cultural unity, historical tradition
and a sign of the faithfulness of the Catalan people to their own specific culture [. . .]'.
At the same time this law enlightened three major aspects of Catalonia's new policy:
1. The position of Catalan is precarious, so Catalonia needs an ambitious language and culture policy at school, of course, but
also in the

< previous page page_194 next page >


< previous page page_195 next page >
Page 195
administration, for civil servants, lawyers, on the radio, television and in the arts (theatre, film production, no subtitles,
translation).
2. This law insists on concepts such as participation of all citizens, democracy, peace, variety of cultures and languages in
Spain and it considers multilingualism and multiculturalism not as a handicap but as a principal benefit for Spain. This leads to
the third important aspect of this law.
3. Catalonia also recognises Spanish-Castillano as the official language, which means in fact that the Generalitat has to
organise the promotion of two main languages and to maintain some local dialects in Aran and Valencia.
This does not mean that Catalonia is interested only in its past. The Generalitat also wants to promote the study of at least two
foreign languages in secondary schools. What are the results of this law? Has it been possible to reach these goals in the past
five years?
It has first to be noted that we have to deal with a situation which is totally different, for example, from Belgium, where the
language acts were promulgated before the Flemish and Wallonian regional official authorities were set up. The existence of the
Generalitat and this law allowed the creation of a department for language affairs which was able to develop a form of language
strategy:
First observation: Such macrosocial projects cannot achieve their aims in five or ten years. They are long-term programmes.
Catalonia is not only a region where Catalan people live. Many people from other parts of Spain, immigrants from Latin
America and the fact that there are socioeconomic differences and different levels of language consciousness made it impossible
to achieve the aim of generalised bilingualism with just one law, one global educational approach and one school model.
Second observation: Catalonia cannot achieve these aims overnight, as there are not sufficient teachers and qualified people able
to carry out this language policy in such a short time.
Third observation: It is not very easy to convince 'Castillano' speaking people coming from other regions in Spain that they also
have to learn Catalan. One has in mind a certain idea of one's own country. The unconscious cultural perception of Spain as one
country with one government, one capital, one language and at least with a 'mono-identity' has to be transformed. All of a
sudden, Spanish embassies abroad have to promote a 'multicountry' which is not obvious.
Fourth observation: It is understandable that a language policy on a macrosocial levelbut surely also at the level of one school
and its environmenthas to be negotiated with the people concerned, and supported by a cultural policy which should clearly
outline the new affective, non-verbalised, identity and culture at which it is aimed.
Since Charlotte Hoffmann proposed an analysis of the role schools should play to reach their bilingual goals in Catalonia, I do
not want to go into

< previous page page_195 next page >


< previous page page_196 next page >
Page 196
further details on this subject. I just want to point out the interdependence between all school patterns and the four general
problems I have mentioned when bilingualism is to be dealt with.
One conclusion of this section could be that it is understandable that many local bilingual experiments failed. Educated in
monolingual schools in mostly monolingual countries, we do not sufficiently realise the importance of formal education as a
mirror of one specific identity and culture. In bilingual education more than in monolingual education there are specific
communication strategies to be created between the political wishes, the local pressure groups and the linguistic and cultural
needs. Those strategies should concern the pedagogical level as well as all extra-curricular activities and social events organised
by the schools.

New Attitudes of Monolingual Majorities: The Case of Italy


Before dealing with this matter, it is necessary to present those minorities that have partly the same historical background as the
national minorities in Spain. Italy is also a country in which many small language communities are living. On the one hand,
there are Albanian, Slovenian, Greek and some other groups who originally came from abroad. On the other hand, it is not
acceptable to state that the Italian population speaks only one form of Italian from Sardinia to Naples or Venice. So, if we want
to understand what is happening now with the monolingual majority, we should not forget to place this new evolution in the
context of a linguistically 'heterogeneous' and 'isolated' country: no other nation speaks Italian and the unity of Italy is very
recent.
This partly explains the situation of South Tirol and, of course, the bilingual educational systemFrench/Italianlegally
institutionalised in Aosta in 1948 with a special autonomy statute. Although this statute gives equal rights and status for both
languages, the reality in the field was and still is different.
The law says that French should be taught as many hours each week as Italian and that some disciplines could be organised in
French. In other words, the language at school is Italian. Local inspectors of primary education explain that this policy was
based on an official recognition and acceptance of a linguistic reality. But since Val d'Aosta is also a region in which people
speak Franco-Provencal besides Italian and French and in which economic independence from the rest of Italy is obviously
growing, the position of the French language becomes more and more precarious.
Since the early 1960s, the whole linguistic situation of Aosta and the attitude of the Italian speaking majority in Italy began to
change. The Treaty of Rome, the participation of Italy in the Common Market, the increasing economic integration in
Europeand at the same time the results of all

< previous page page_196 next page >


< previous page page_197 next page >
Page 197
the research made on the field of applied linguistics and language teachinghad some practical influence on public opinion.
Titone and foreign researchers like Penfield, Roberts and Balkanstated at that time that early bilingualism has no negative
consequences on the child's general development. Positive psychopedagogical requirements and conditions in favour of bilingual
education are described. Experiences prove that this type of education should in the early stages of the children's development
have a less formal, more natural touch than traditional language teaching at the secondary level. The importance of the parents'
attitudes and the use of the second language not only as a subject, but also as a medium of teaching and communication, are
emphasised. Last but not least, the need of specific teacher training is acknowledged.
What were the consequences of these new facts? In Aosta at the end of the 1970's the local authorities started with a new
concept of bilingual education, language planning and teacher training. A totally new method elaborated by the French
University of Grenoble for Kindergarten (age three) has been introduced in French/Italian on a part-time basis (50% French;
50% Italian) everywhere in the Aosta Valley and the 'threshold level' methods of the European Council are still tried out at the
secondary level.
In the rest of Italy experimental programmes were organised in all the main cities such as Venice, Rome, Milan, Torino, mainly
the support of the local authorities. All these projectsthe most important, I.L.L.S.E., started in 1977 with the financial support
and the guidance of the Italian Ministry of Public Instructionare based on the fundamental statement that it is possible to reach a
certain level of multilingualism in French, German and English if the teachers are trained and if the children are taught at least
three to five hours a week in the foreign language at primary school level.
Even if this statement has not yet been absolutely assessed, last year the results of these experiments led to a great decision in
the Italian Parliament: the introduction of one foreign language for three or four hours each week for children (nine or ten) in all
primary schools. This means that nearly 65,000 teachers need special teacher training before 1989, when this law comes into
force. This also means a number of problems which are not yet solved. However, the phenomenon as such is remarkable for the
attitude of an old South European country.

From the Integration of Immigrants' Children in France to the Linguistic Integration of France in Europe
The linguistic situation of France is quite complicated and there are as many prejudices in France as elsewhere in the approach
to language policy and in public opinion. The French educational system is historically based

< previous page page_197 next page >


< previous page page_198 next page >
Page 198
on the linguistic unification of the country. It should not be forgotten that for generations French primary teachers found it
necessary to fight against seven 'regional' or 'national' languages. There was no future, no social promotion for the farmers'
children who could not speak French when they had to leave their homes to find work in the major French cities. Equality and
democracy were synonymous with monolingual education in French. In Brittany, one could read on the school walls 'It is
forbidden to spit and to speak Breton'.
This attitude in language matters can also be explained because France had colonial territories all over the world and because
French used to be very popular in all foreign elites. The sentence, 'French is the most beautiful language in the world' was
commonplace in the 1950's . . . not only in France. Even after 1945, nearly everybody in Western Europe was monolingual and
during the conferences organised in the early 1950's by the Council of Europefor example in Wiesbaden (1954)famous experts
explained very seriously with words such as dislexia and schizophrenia how dangerous it is to become bilingual. These ideas
have not yet completely disappeared, but it can be said that, in France, four major trends have contributed to change this
situation.
One one side, since the colonial territories have become independent and because France accepted important minorities from
North Africa, Portugal, Spain, Italy, etc., the French Ministry of Education has agreed to provide special courses called
'Language and Culture' for immigrants' children. These courses are organised with the direct help of the foreign governments
concerned, who send and pay the native language teachers. The French Ministry of Education provides separate 'French as a
foreign language' classes. These classes are called CLIN and CLADI' and 'A' stand respectively for 'integration' (at primary
level) and 'adaptation' (at secondary level). This system for better language proficiency in French can be compared in its
evolution with the model used in the United Kingdom.
In the middle of the 1970's, the French Ministry of Education tried to meet the wishes of the foreign families, teachers and
governments in setting up mother tongue courses at primary level (age 6-11) besides the normal French-only curricula, late in
the afternoon, on Wednesdays or on Saturdays when these children were free. French children were not accepted in these
courses and, as they were given for three hours a week, it is difficult to say that these courses are aimed at anything other than
maintaining a transitional contact with family culture and language.
The goal of this system was to facilitate the integration and assimilation of foreign children who had school problems. This
model still exists but new models were created at the end of the 1970's and at the beginning of the 1980's. In order to understand
why this model has been partly changed, it is important to mention the other new factors which had a major influence on
politicians and public opinion.

< previous page page_198 next page >


< previous page page_199 next page >
Page 199
The second factor is the ideological meaning of the research carried out on the problem of human knowledge and indirectly on
the problem of bilingual education. Scientists such as the linguists, Martinet, Tabouret-Keller, Daben and Jacquart and
Changeux in their research on genetics and the constitution of human brain, had a certain influence on the teachers' world. They
all proved the importance of education in general for children aged 1-6, and the effectiveness of an early bilingual-planned
education.
The third important factor or tendency has been a deep evolution in French public opinion since the end of the 1970's. Like
Italian people, French public opinion seems to find it more and more necessary to learn foreign languages at a quite early age.
Here are some results of a survey made in 1979. Sixty-six per cent of the parents found that starting as today in France, with a
first foreign language at secondary level at the age of 12, is too late and 56% of them thought that three or four hours a week
were not enough. At the secondary level, it is possible to choose theoretically between 13 foreign languages (!) but most parents
prefer English (93%), German (62%), Spanish (21%), or Italian (5%). Sixteen percent of the parents preferred the introduction
of a first foreign language before the age of 6, 16% at 6, 8% at the age of 7, 14% at the age of 8, 8% at the age of 9, 14% at the
age of 10, 12% at the age of 11. Only 9% of the parents agreedin 1979with the normal French regulations on that subject.
The last important evolution is the identity reactions of the language minorities in several parts of France. Since the Council of
Europe has supported several communities and minorities of lesser used languages and since the language policy has changed in
Spain, the French Basque, Occitan, Breton and Catalan communities have organised some private schools in the hope that the
French Ministry of Education agrees to finance their bilingual kindergartens and primary school model as soon as possible.
These three communities have nearly the same model. At kindergarten level (age 3-6) and in the first form of the primary
school, the minority language is the only language to be used. These children learn how to read and write in Basque or Catalan
with the help of books and methods mostly made and bought in Spain. In Brittany, there are regular contacts between the
'Diwan' schools and Wales. From the second form on (age 7-8), French is introduced during six hours a week. At the end of the
primary school (age 10-11), since there is only one secondary school with an important part in Breton, the pupils are following
the normal French curriculum with one foreign language.
All these four important facts have a combined effect on the attitude of the Ministry of Education at the moment when
'decentralisation' of the educational policy cannot be avoided and all the more so that everybody in France is now speaking
about the European integration planned for 1992 and about the economic necessity to imagine an increasing number of projects
like Eureka, First, Spirit and Erasmus on a European scale. 1

< previous page page_199 next page >


< previous page page_200 next page >
Page 200
What are the observations to be made now in relation with these four factors in our educational system?
Firstly, in France 10 or 11% of all the children (age 3-11) are citizens of a foreign country, which means 650,000 children:
200,000 from Algeria, 160,000 from Portugal, 90,000 from Morocco, 35,000 from Tunisia, 35,000 from Spain, 30,000 from
Italy, 20,000 from Laos and Cambodia, 30,000 from Turkey and 20,000 from other countries in Africa. In addition to the model
I have already described, three other models for mother tongue teaching appeared:
Second Model: Three hours each week with the immigrants' children withdrawn from their class during the ordinary school
time; that is to say, when the other pupils are continuing their work. Cultural activities related to the original, native country are
often organised in the evening or during the week-end. This model is also variable and depends on the staff of the school.
Sometimes French children can be accepted in these classes considered as 'mother tongue and culture' courses.
Third model: this model is actually not a model. It corresponds to the situation when a foreign pupil arrives in a school in which
nothing has been planned for his integration.
Fourth model: This model is the model created in the north of France. French and immigrants' children, from Portugal or
Algeria or Morocco or Italy, are put together in the same class. They all get about six hours a week in that foreignor
nativelanguage. All subjects can be taught in that language as far as the teachers' work is related to the official French syllabus.
These classes have contacts with schools in the relevant foreign country. If possible, we organisein the fourth and fifth form (age
10-11)a stay of two or three weeks in that particular twin-school in the foreign country.
The French Ministry and some foreign countries, such as Italy and Portugal, are considering some extension of this fourth
model. We have proved (LePong, 1986) that immigrants' children have better results even in French when they have six hours in
their native language together with the French children. We also proved:
1. that the French children in these special classeseven if they get 21 hours in French instead of 27have results the same as or
better than other French children in 'ordinary' classes. This, of course, means some teacher training and co-ordination between
the foreign and French teachers.
2. that after five years these French children, at the secondary level, are able, with not too many difficulties, to follow a lesson
given in history or geography in that foreign language.
The second main evolution we can consider is the influence that the Basque, Breton and Catalan schools have on the ordinary
primary and secondary French schools in the concerned regions. It is now possible to be taught (three hours a week) in those
minority languages on a voluntary

< previous page page_200 next page >


< previous page page_201 next page >
Page 201
basis, if there are enough pupils attending those lessons.
The third important trend and the most fundamental one, is what is happening to the majority, the monolingual children in
France in this new context. Since the early 1960's there have been some experimental kindergarten and primary schools with
different models of second language teaching in English, German, Spanish and Italian. Nowadays, the pressure from French
parents for early foreign language teaching on the local educational authorities is more and more effective.
In 1979-80, 3,210 teachers, 1,600 schools, 4,200 classes were officially involved in the 'early' introduction of a foreign language.
This number does not mean a revolution in our educational system, but it is increasing in the public system and also in private
'English mini-schools'. These mini-schoolsand 'pop-gartens'(age 3-6) are organised during the week-ends or on Wednesdays.
There are more than 1,200 of these initiatives in France. In the project in Northern France, we organised second language
teaching in English, German and Dutch on the same basissix hours per weekas in Portuguese, Arabic and Italian. All the
involved schools24 todayand the 120 teachers from six countries are working together in the same project. Nearly every day a
new primary school asks for the introduction of a foreign language. For the secondary level, the aim is to plan the teaching of
language, literature, history, geography and even more subjects if possible in all these six languages up to university level. At
this level, the attitude of the teachers and the parents is generally very positive.
These trends, these facts, all this applied research could give the impression that France has changed overnight. This would
obviously be a mistake. It is just a beginning. A very small percentage of the majority group is now involved in those projects.
Traditions and habits cannot change that fast. Our problem is somewhat the same as in Catalonia. We need time and we have to
solve a practical problem: how can we reach a regularly increasing number of pupils involved in early bilingual education and
cover the whole country? What languages have to be taught? At the moment the local educational, economic and political
authorities are discussing a new strategy based on three phases: development of the efficient third model, motivation for new
steps through communication policy, and extension. These three phases also include programmes of teacher exchanges at EEC
level, international twin-school projects and travelling. The use of computers for all school levels in foreign and in mother
tongues is planned for French as well as for immigrants' children.
The conclusion which could be drawn from the general evolution in Spain, Italy and France seems to be obvious. Local, regional
and foreign languages are gaining in importance in Southern Europe. Here and there, the presence of immigrants' children starts
to be considered not as a handicap but as a useful and enriching fact for our old European societies, especially because most of
the migrants are from European origins. At the same time, these contacts with different cultures give us a chance to strengthen
our

< previous page page_201 next page >


< previous page page_202 next page >
Page 202
democracies. However, this can only be possible if the professions involved and public opinion are able to manage this form of
mental and social revolution.
Am I an optimist? The fact is that the changes in economy all over the world and the demographic features speak for
themselves. What will happen to the occidental world in the next 50 years if there is no solution to this language and cultural
challenge? We should not forget that there are only about 150 countries in the world. In those countries, almost 5,000 official
languages are spoken and, in the present world, monolingual societies have a demography much less important than multilingual
societies.

Notes
1. Names of European projects concerning: a) communication; b) use of new technologies; c) common standardisation of
computer-programs and languages; d) exchange of student integration of study curricula, higher education networks.

References
Changeux, J.P. (1985) L'Homme Neuronal. Paris: C. Levy.
Daben, L. (1986) In actes du colloque. Les Enseignements bilingues en Europe. Lycée Francais de Madrid. Madrid 1986. pp. 27-
29.
Hoffmann, C. (1987) 2 Sprog 2 Kulturer, Abstracts. p. 53.
Jacquart, A. (1983) Introduction. In R. Cohen (ed.), Plaidoyer pour les apprentissages précoces. Paris: Seull.
LePong (1986) Résultats d'enquête concernant les sections internationales franco-italiennes et franco-portugaises de la
circouscription de Tourcoing I. Rapport à l'inspection académique. Mai 1986.
Martinet, J. (1986) Colloque sur le Bilinguisme en Europe. Landerneau. Organisé par Diwan.
Tabouret-Keller. M. 1986) Intervention. Actes du Colloque. Les Enseignements bilingues en Europe. Lycée Francais de Madrid.
Madrid 1986. pp. 20-27.

< previous page page_202 next page >


< previous page page_203 next page >
Page 203

Bilingualism, Education and Identity


John Edwards
Department of Psychology, St Francis Xavier University
Antigonish, Nova Scotia, Canada B2G ICO.
Abstract. What is the role of the school in supporting minority group identity, particularly through programmes of
language maintenance? It is argued in this paper that schools acting in isolation can do little in this regard, although they
can and should promote tolerance for cultural and linguistic diversity. The paper begins with a definition of ethnic
identity, in which it is stressed that no single objective marker is necessary for the continuation of identity. Next, the
relationship between language and group identity is discussed and, following this, I turn to consider education itself.
While language shift is, in many circumstances, seen to have an inevitability which the school cannot hope to counter, it is
also pointed out that identity can be maintained through periods of language shift.
I take it as a working assumption that bilingualism is a good thing per se, and should not be seen as a 'problem'. It is important
to note at the outset, however, that a distinction should be drawn between stable and unstable bilingualism. This rests upon the
continuation of viable domains for each languageif these exist, then clearly bilingualism can be an enduring phenomenon. But if
over the course of time a language loses exclusivity within a domain, and one (other) language comes to serve across all, then an
unstable bilingualism results which does not bode well for the continued survival within the group of both languages. The
simple rule seems to be that people will not indefinitely maintain two languages where one will serve all their needs. This, then,
is a scenario for language shift. However, I shall be arguing below that such shift does not necessarily mean the loss of group
identity. So, this paper is essentially about conditions of unstable bilingualism and their implications for identity.

< previous page page_203 next page >


< previous page page_204 next page >
Page 204
I also assume that bilingual education can be seen as a useful and necessary innovation. This is particularly so in its so-called
'transitional' format. On the other hand, 'maintenance' varieties of bilingual education may pose certain difficulties when they are
seen as servants to the cause of cultural pluralism or the support of group identities seen to be at risk (see below).
This paper, then, briefly examines some aspects of the relationships among the three terms given in the title. More specifically, I
intend to consider the connection between the maintenance of an ethnic group's original language and the continuity of group
identity. It then becomes necessary to bring education into the arena, since school programmes of one sort or anotherparticularly
those dealing with minority languageshave traditionally been looked to as supporters of identities seen to be at risk in some
larger environment; that is, the implications of some programmes go some way beyond purely pedagogical objectives. The
maintenance of original group language is commonly seen to be central for identity continuity. Equally, as I have just noted,
school is commonly seen to be a potential bulwark for identity. This paper discusses both of these perceptions, and then
considers some of the implications of this discussion.

Ethnic Identity
The first step here is to attempt to formulate a definition of ethnic identity. There have, of course, been manyalthough Isajiw
(1980), in examining 65 studies of ethnicity, found that 52 of them gave no explicit definition at all. Bearing in mind Isajiw's
observations, Edwards (1985) presented a definition of ethnic identity which took four major points into consideration: (1)
ethnic identity need not be a minority phenomenon; (2) perceived group boundaries can continue across generations even though
the cultural content within these boundaries may have changed radically (see Barth, 1969); (3) objective trait descriptions do not
fully encompass the phenomenona sense of 'groupness' may be very important here (see data from Nova Scotia, reported in
Edwards, in press; Edwards and Chisholm, 1987; Edwards and Doucette, 1987); (4) so-called 'symbolic' ethnicity (see Gans,
1979) should be taken into account. Edwards' definition is thus:
Ethnic identity is allegiance to a grouplarge or small, socially dominant or subordinatewith 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 sustained by shared objective characteristics (language, religion, etc.), or by
more subjective contributions to a sense of groupness, or by some combination of both. Symbolic or subjective
attachments must relate, at however distant a remove, to an observably real past. (Edwards, 1985: 10)
The major import of this definition for present purposes is its implication that no particular objective marker is necessary for a
continuing group

< previous page page_204 next page >


< previous page page_205 next page >
Page 205
identity. Thus, indigenous groups (for example, the Irish) and immigrant groups such as those found in North America may
retain a sense of cultural boundaries while the content within these boundaries (for example, language in its ordinary,
communicative sense) has altered dramatically over the generations.

Language and Groupness


If it is true that the original group language is not an essential component of identity, it is equally true that many yet consider it
to be essential. For example, the German romanticism of the early nineteenth century made the assumption that language was
the central pillar of nationalistic group feeling. Thus, Johann Herder observed that
even the smallest of nations . . . cherishes in and through its language the history, poetry and songs about the great deeds
of its forefathers. The language is its collective treasure. (Barnard, 1969: 165)
Similarly, Wilhelm von Humboldt insisted that language was the 'spiritual exhalation' of the nation (Cowan, 1963: 277). Johann
Fichte noted, in his famous Addresses to the German Nation (1968, original 1807), that because the German language was
superior to others, the German nation was superior. Thus, we can easily trace the language-group relationship from the time of
these German romantics (this is not to say, of course, that sentiments of a similar kind cannot be found still earlier in history).
Smith (1971: 182) summarised things rather neatly when he observed that 'the notion that nations are really language groups . . .
derives from Herder's influence'. In his most recent book, incidentally, Smith (1986) rejects the idea that language need be the
main differentiating mark of ethnicity.
I think it is clear enough why language should have this centrality for manyit is a highly visible marker in which communication
and sentiment may be tightly bound. So, efforts are often mounted to support and encourage languages seen to be at risk, as a
way of maintaining the group coherence of their speakers. Still, it can be argued that language shift occurs because of contact
between groups, and usually has an underlying economic basis; language shift is a symptom of such contact. It follows that
attending to language alone is highly unlikely to reverse or significantly alter the larger forces set in motion by group contact.
It is useful here to consider a distinction between language in its communicative sense and language as a group symbol. For
majority-group speakers in a mainstream setting, both of these aspects remain important and connected. For minority groups,
however, communicative language shift may occur while leaving behind a symbolic value of importance. For example, the 1975
report of the Committee on Irish Language Attitudes Research indicated that while few spoke Irish regularly, and few were
interested in organisations

< previous page page_205 next page >


< previous page page_206 next page >
Page 206
aiming to resuscitate it, most felt that Irish had a special claim on their affections, thus retaining some sort of group-marker
status (see also the comments by Eastman (1984) on 'associated languages'). Ignorance of this communicative-symbolic
distinction may lead to quite inappropriate interventions on behalf of groups for whom the split has occurred.

Education and the Role of the School


Many of those who see education as a handmaiden to group and group-language maintenance or as a servant to the cause of
pluralism would appear to implicitly support Kedourie's statement (1961: 83-4) that
on nationalist theory . . . the purpose of education is not to transmit knowledge, traditional wisdom, and the ways devised
by a society for attending to the common concerns; its purpose rather is wholly political, to bend the will of the young to
the will of the nation.
Ravitch (1981) has observed, of American education, that schools have been increasingly burdened with matters of social policy,
that they run the risk of becoming 'sociological cookie cutters'. Of course, all education is political, but presumably there are
degrees. Bullivant (1981) touches upon the topic when, in a study of six plural societies, he discusses the dilemma of 'civism'
and 'pluralism' at school. He is fearful of a weakening of the school's transmission of essential common elements; a general-
knowledge core curriculum is what minority students most need in order to succeed in the wider society. Relatedly, Bullivant
notes that formal attention to purely educational programmes may distract from the solution of real socio-political problems
faced by minority groups. Others (for example, Breton, Reitz & Valentine, 1980) have noted how governments may adopt a
pluralistic stance-particularly via school programmesas a way of keeping minorities subordinated while appearing to attend to
their needs.
The major problem in using the educational setting to bolster minority-group language and identity is that school programmes
are typically put in place to assist groups seen to be at risk. Given that a complex of social factors has produced this very
situation, it is apparent that, very often, the school is essentially being asked to counteract, or even to reverse, larger societal
dynamics. There is little evidence that schools, which tend to reflect the larger society rather than to lead it, can significantly
affect minority-group language and identity. There can be, in other words, too much emphasis upon what schools alone can do,
given that their efforts are often dwarfed by social pressures outside their gates.
A useful and informative case study of education, language and identity is the Irish. Between 1600 and 1800, English made
steady advances and by the time the National School system was established in 1831, the number of Irish monolinguals was
very small. Bilingualism, in this context, was usually and typically a way-station on the road to English monolingualism.

< previous page page_206 next page >


< previous page page_207 next page >
Page 207
It may be, then, that the Irish case resembles that of Manx, of which O'Rahilly (1932: 121) observed: 'When a language
surrenders itself to foreign idiom, and when all its speakers become bilingual, the penalty is death.'
Certainly, it has been noted by many that the English-run National School system in Ireland was, with other factors, a
significant cause of the decline of Irish; this so-called 'murder machine' (Pearse, 1976, original 1912) excluded Irish from the
curriculum. Yet it must be remembered that a massive language shift to English was continuing outside the school gates, and
that the population generally acquiesced in this shift. Thus, de Fréine (1977: 84) noted that 'most of the reasons adduced for the
suppression of the Irish language are not so much reasons as consequences of the decision to give up the language'. Similarly,
O'Brien (cited by Ó Conaire, 1973: 125) observed that 'the present extremity of the Irish language is due mainly to the fact that
the Gaels deliberately flung that instrument of beauty and precision from them' ('flung' may be rather too strong a word here).
With the founding of the Free State, Irish was enshrined as the national and first official language. By this time, of course, the
mass of the population had long since switched to English. In these circumstances it seems obvious that educational programmes
alone would have a very hard time attempting restoration. Yet the government essentially passed this burden to the schools.
From the beginning, Irish was to be a compulsory subject, and infants' classes were to be conducted entirely through Irish-even
though this was an unknown language for the great majority. These measures arose from recommendations of the Irish National
Teachers' Organization in 1921. Since that time, teachers have increasingly questioned the implicit decision of the government
to revive Irish through schools. One teacher (Comber, 1960: 27) noted that 'the teacher loses heart flogging a dead horse while
the experts debate whether another whip might not revive him'. Another (Harrison, 1976: 35) observed that 'Irish . . . is on its
way out. Don't blame the teachers for its demise.' Although the school system in Ireland has, over the last 65 years, given
everyone a very thin wash of Irish competence, it has, unsurprisingly, not effected a general restoration of the language.
Most recently, there seems to have occurred a realistic appreciation of the concerns of teachers and others. A government White
Paper of 1980 reaffirmed the official aim of Irish restoration, but recognised that schools should not be expected to achieve this
on their own. Given the history of Irish and English in Ireland, as well as the current situation, it would seem that extra-
educational dynamics are not favourable to a revival of Irish on a broad scale; consequently, such a revival is extremely
unlikely. Nevertheless, as noted above, the Irish language does appear to retain a symbolic status in the perceptions of many
Irish people. Communicatively, however, the Irish identityand it is a strong and distinctive oneis expressed through English.
Another useful perspective on language and education is provided by the bilingual education movement in the United States. It
has proved to be a

< previous page page_207 next page >


< previous page page_208 next page >
Page 208
controversial area, and much of the debate is over the so-called 'transitional-maintenance' distinction (to which I alluded at the
beginning). That is, schools can provide early education in the mother tongue of non-English-speaking children until their
facility in English allows them to be 'mainstreamed'; on the other hand, as the name implies, maintenance varieties are seen as
more permanent fixtures in which bilingual instruction continues throughout the school career or at least a substantial part of it.
The early thrust of the American Bilingual Education Act of 1968 was compensatory in nature, was not meant to overturn the
melting pot or to institutionalise pluralism, and led in the main to transitional programmes. However, much of the academic
community concerned with bilingual education is opposed to a transitional philosophy, the commonest view here being that
transition actually expedites assimilation, which is itself regarded as a bad thing (it is worthwhile pointing out here, however,
that assimilation may not always be an undiluted evil, at least not in all its aspects; see Edwards, 1985). Yet most American
bilingual education remains of a transitional nature, and there are grounds for thinking that this accords with the desires of those
most directly concerned. Fishman (1980) has contrasted the 'numerous' and 'tragically destructive' transitional programmes with
the better ones sponsored by ethnic communities themselves, these being few in number and 'weak'. In asking ourselves why
this situation exists, we should recall the decline of 'ethnic' schools in the last century and the early years of this one. Put
simply, this decline occurred because the communities concerned no longer felt the need for such schooling, and were anxious
for their children to move firmly into the American mainstream, thereby making the upheaval of emigration psychologically and
materially worthwhile. Schlossman (1983a, 1983b) has looked at early German and Spanish schools in the United States, and
has concluded that the ethnic communities they served never considered public schools to be potential preservers of identity.
While transitional language programmes seem reasonable and deserving of public support, maintenance education is less easy to
defend from an all-society point of view. It involves direct administrative involvement in identity retention and reflects the idea
that diversity is not only to be approved of and tolerated but is to be promoted to the level of official policy. There is no
evidence to suggest that meaningful aspects of ethnicity can be held in place by outside interventionagain, one must think of the
strong unofficial pressures which often militate against ethnic-marker retentionmuch less ones which are visible distinctions
(like language). There is little hope, in any event, that schools alone can provide such intervention successfully. Finally, there is
little evidence that such intervention is actively supported (passive tolerance and lip-service are another matter) by large
segments of the population, minority or majority (see, for example, Edwards & Chisholm 1987; Edwards & Doucette, 1987).
If school language programmes cannot, by themselves, maintain ethnic identity seen to be at risk, what is an appropriate stance
for schools in

< previous page page_208 next page >


< previous page page_209 next page >
Page 209
societies where diversity is the rule and where there is at least tolerance or passive support for the continuation of minority-
group identity? Schools can, above all, provide tolerant atmospheres which do not act against the expression and continuation of
groupness. They may not be able to actively support group identities but they can reflect the heterogeneous society of which
they are a part. Schools, I believe, must continue to offer to all pupils a solid grounding in core knowledge and skillssomething
which minority-group parents are usually greatly in favour of (see, for example, Stone, 1981).
With regard to the provision of so-called multicultural education, my position is that all education worthy of the name is
multicultural. It should be part and parcel of education generally to show an awareness of diverse cultures and to develop an
appreciation of human difference. But, the schools' role is not primarily one of cultural transmissionalthough some of this is, of
course, inevitable. Schools can lead society in the promotion of tolerance and understanding, and to abdicate responsibility here
would be to renege on the basic goals of education itself. To attempt more active intervention, to engage in programmes
specially planned to bolster minority-group identitythis may be neither desirable nor practical.

Conclusions
The overall message concerning language, education and minority-group identity is at once negative and positive. There is, on
the one hand, a general inevitability to language shift under certain conditions and no amount of educational support can
significantly affect the powerful social currents which produce this shift. It is probably better, however, to speak of language
change under conditions of contact rather than language loss (as is commonly done), since alteration of certain markers of
identity accords better with social dynamics. There is, after all, never a question of loss, pure and simple, with nothing to
replace the abandoned communicative language; another language gradually comes to serve and, in the transitional period,
bilingualism is the usual bridge. Bilingualism in such circumstances is usually, then, of the unstable variety and represents an
instrumental expansion of the linguistic repertoire in response to altered environments; it need not entail any sort of
'schizophrenic' identity pattern.
However, if language shift is the negative side of the coin, the positive is that minority-group identity can be, and commonly is,
maintained through and beyond the transitional times made inevitable by social evolution. Groupness, it would seem, is a
tenacious quantity and is capable of surviving changes in any objective marker, including language.
Language revival or maintenance, cultural pluralism, stable bilingualismthese cannot be instituted by fiat. Neither should we
expect schools alone to successfully counter strong social tides. If language is seen

< previous page page_209 next page >


< previous page page_210 next page >
Page 210
to be at risk, it is often because of a finely meshed social evolution. To remove it from risk would entail wholesale reworking of
history, a broad reweaving of the social fabric; this has not usually been considered a practical exercise.

References
Barnard, F. (ed.) (1969) J.G. Herder on Social and Political Culture. Cambridge: Cambridge University Press.
Barth, F. (ed) (1969) Ethnic Groups and Boundaries. Boston: Little, Brown.
Breton, R., Reitz, J. and Valentine, V. (1980) Cultural Boundaries and the Cohesion of Canada. Montreal: Institute for
Research on Public Policy.
Bullivant, B. (1981) The Pluralist Dilemma in Education. Sydney: Allen & Unwin.
Comber, T. (1960) The revival. Múinteoir Naisiunta, 5(7), 23-7.
Committee on Irish Language Attitudes Research (1975) Report. Dublin: Government Stationery Office.
Cowan, M. (1963) Humanist without Portfolio: An Anthology of the Writings of Wilhelm von Humboldt. Detroit: Wayne State
University Press.
De Fréine, S. (1977) The dominance of the English language in the nineteenth century. In D.Ó Muirithe (ed), The English
Language in Ireland. Cork: Mercier.
Eastman, C. (1984) Language, ethnic identity and change. In J. Edwards (ed.), Linguistic Minorities, Policies and Pluralism.
London: Academic Press.
Edwards, J. (1985) Language, Society and Identity. Oxford: Basil Blackwell.
(in press) Gaelic in Nova Scotia. In C. Williams (ed.), Language in Geographic Context. Clevedon, Avon: Multilingual Matters.
Edwards, J. and Chisholm, J. (in press) Language, multiculturalism and identity: a Canadian study. Journal of Multilingual and
Multicultural Development, 8, 391-408.
Edwards, J. and Doucette, L. (1987) Ethnic salience, identity and symbolic ethnicity. Canadian Ethnic Studies, 19, 52-62.
Fichte, J. (1968) Addresses to the German Nation. New York: Harper & Row.
Fishman, J. (1980) Language maintenance. In S. Thernstrom, A. Orlov & O. Handlin (eds), Harvard Encyclopedia of American
Ethnic Groups. Cambridge: Harvard University Press.
Gans, H. (1979) Symbolic ethnicity: The future of ethnic groups and cultures in America. Ethnic and Racial Studies, 2, 1-20.
Harrison, M. (1976) The revival of Irish. Secondary Teacher, 6(1), 34-5.
Ireland (1980) White Paper on Educational Development. Dublin: Government Stationery Office.
Isajiw, W. (1980) Definitions of ethnicity. In J. Goldstein and R. Bienvenue (eds), Ethnicity and Ethnic Relations in Canada.
Toronto: Butterworth.
Kedourie, E. (1961) Nationalism. London: Hutchinson.
Ó Conaire, B. (1973) Flann O'Brien, 'An Béal Bocht' and other Irish matters. Irish University Review, 3, 121-40.
O'Rahilly, T. (1932) Irish Dialects Past and Present. Dublin: Institute for Advanced Studies.
Pearse, P. (1976) The Murder Machine and Other Essays. Cork: Mercier.
Ravitch, D. (1981) Forgetting the questions: The problem of educational reform. American Scholar, 50, 329-40.
Schlossman, S. (1983a) Is there an American tradition of bilingual education? German in the public elementary schools, 1840-
1919. American Journal of Education, 91, 139-86.
(1983b) Self-evident remedy? George I. Sanchez, segregation, and enduring dilemmas in bilingual education. Teachers College
Record, 84, 871-907.
Smith, A. (1971) Theories of Nationalism. London: Duckworth.
(1986) The Ethnic Origins of Nations. Oxford: Basil Blackwell.
Stone, M. (1981) The Education of the Black Child in Britain: The Myth of Multiracial Education. London: Fontana.

< previous page page_210 next page >


< previous page page_211 next page >
Page 211

Where Do We Go From Here?


Concluding Panel Discussion

Gabriele Kasper: The need for an Interdisciplinary Research Strategy in Second Language Studies
I would like to make some brief remarks about what I see as a necessary research strategy in second language (L2) studies that
have a practical orientation towards bilingual education. My suggestion is in no way original, but rather what seems to me a
timely reminder.
Many of the topics discussed in the last few days will be of central concern for researchers and educationalists during the
coming years: learners' strategies, discourse procedures in cross-cultural communication, classroom interaction, the inter-relation
of language, culture and identity, bilingualism and cognitive development. The question I would like to put forward is how these
and other topics, whose relevance to bilingual education is uncontroversial, can be most appropriately investigated.
Let me do this by way of an example from this conference. You may remember that in the discussion of learning strategies
following Dr Chamot's main plenary paper, Dr O'Malley remarked that the trouble with current theories of L2 learningUniversal
Grammar, the Monitor Model, the Interactional Approachwas that they lacked a psychological basis in cognitive theory. Indeed,
an interesting theory of L2 learning has to explain how L2 information is selected from input, processed, stored, and accessed in
real time. The model presented by Dr O'Malley (this volume, p.43) takes care of the processing dimension. Yet, as is typical of
many models of cognitive processing, it does not take into account the specificity of the information to be learned and used, that
is, the structure of linguistic information. As long as we assume, as linguists and psycholinguists do, that the structural and
functional properties of linguistic knowledge are qualitatively different from, say, motor skills or mathematical knowledge, our
models of L2 learning and use have to account for such properties. I understand from Dr O'Malley that such a theory has
recently been suggested (MacWhinney & Anderson, 1986). Furthermore, as L2 learning presupposes already existing L1
knowledge, a theory of L2 learning and use will have to incorporate prior linguistic knowledge and its interaction with the
learners' interlanguage knowledge. I am not sure whether the new model is readily applicable to interlanguage knowledge.

< previous page page_211 next page >


< previous page page_212 next page >
Page 212
But even if it could, this would not be the whole story. Let us assume we would like to establish a framework for the analysis of
learning strategies. Such a framework would at least have to accommodate the following components:

In order to describe and explain the impact of these components on the use of learning strategies, and the interaction of learning
strategies with L2 input and prior knowledge, theories and methodologies from at least the following scientific disciplines have
to be available, in relation to the different components:
Linguistics: the structure of L2 input and previous linguistic knowledge;
Second language research: the interaction between L1 and L2 knowledge; Cognitive psychology: the organisation of other
knowledge, and the operation of learning strategies;
Other branches of psychology (social, developmental; psychoanalysis): affective variables (attitude, motivation; 'character'
in psychoanalytic terms);
Anthropology: sociocultural background.
It follows that the task of developing an analytical and educational framework for learning strategies requires a highly
interdisciplinary approach. Indeed, such an approach is requisite for most research questions in second language studies. In order
to tackle a research task in a problem-oriented way, such theories and methods have to be selected from each of the relevant
disciplines that are assessed as most appropriate for the given task. This implies that the researcher has to be able to:
compare alternative theories and methodologies to each other;
understand the presuppositions and consequences of their differences;
evaluate their adequacy for the task;
and to apply the selected theory and method to the research problem in combination with theories and methods that have been
selected by the same procedure from the other disciplines.

Clearly, with the exception of the happy few individuals who are on lifelong research grants or born geniuses, and preferably
both, this is not a feasible enterprise for one researcher, or even for a group of investigators with the same professional
background, given such vast, diversified and rapidly expanding fields as psychology or linguistics. And yet it is more the rule
than the exception, at least in this country, that one person or a group of people with very similar qualifications have to make
decisions that are crucial for an

< previous page page_212 next page >


< previous page page_213 next page >
Page 213
investigation and its outcome, without having the necessary professional insight.
How are such decisions made, then? If we asked a representative sample of, say, linguists to give us a self-report, a typical
answer could be as follows: 'I choose from the recent literature a procedure that is comprehensible to me as a non-expert, suits
my personal inclinations (i.e. is in accordance with my unscientific preconceptions of reality) and doesn't involve any statistics
demanding a higher degree of sophistication than the chi square'. Can you recognise yourselves in this statement? I can, to my
dismay.
Now while this widely adopted procedure brings about manageable research designs, it hardly results in the most appropriate
ones. In order to achieve adequate decision-making in L2 research, investigators have to collaborate in a problem-oriented way,
so that they can profit from each others' professional insights. I am aware of the communication problems such a collaboration is
likely to entail, due to different conceptualisations and research traditions. However interdisciplinary co-operation seems to me
the only way to secure that L2 research is conducted with the required professionalism.
In many countries, rigid traditional boundaries between different but related fields have been detrimental to interdisciplinary co-
operation. These boundaries have to be overcome. One incitement to enhance researchers' willingness to participate in
interdisciplinary projects would be for them to obtain as much credit for this work as for studies within their traditional
disciplines. Another boost to interdisciplinary co-operationtverfagligt samarbejdecould be given by the funding agencies, in that
they specifically support projects beyond and across the established academic fields.

References
MacWhinney, B. and Anderson J. (1986), The acquisition of grammar. In I. Gopnik & M. Gopnik (eds), From Models to
Modules. Norwood, N.J.: Ablex, 3-25.

Ellen Andenæs: The Researcher's Dilemma, or the Dual Task of the Applied Researcher
The participants in this congress are (mostly) either second language teachers, or researchers, or both. Personally, I feel I belong
to the 'both' groupand I shall address myself to both.
When I first started teaching Norwegian as a second language, I very quickly found out that I was not at all qualified for it. I
knew nothing about my students; about their needs, their backgrounds, or their native languagesnor could I ask them about these
things, since we had no language in common. Since I was a university graduate, I concluded I must have studied the wrong
subjects, and quickly went back to the university to supply myself with the knowledge and information I needed to do my work.
I was sorely disappointed. For some of the languages I needed to know about, a few linguistic reports were availablebut on the
whole, I found

< previous page page_213 next page >


< previous page page_214 next page >
Page 214
them unintelligible, and not very relevant. For most of the questions I asked (such as 'how do I find out what these people need
to know?' and 'how do I teach it to them?') there simply were no answers to be had. I found this rather shockingand set out to do
research myself, hoping to provide at least some of the answers.
In Norway, this is a very common background for second language research. I see this as typical of the field. Typical of the field
also, that teachers realise their need for knowledge, and are extremely willing to turn to 'experts' in order to get advice. (Indeed,
one of the great rewards of working in this field is the feeling that one's work is considered useful and needed.)
So what we have, then, is, on the one hand, teachers who turn to researchers for answers on how to do their work, on the other,
researchers who entered the field for precisely the same reason. This seems like a happy situation. But it is not as idyllic as all
that. More often than not, teachers are not satisfied with what research has to offer. A comment often heard is 'Yes, yes, very
interesting, I'm sure. But how does it help me teach my students?'
This leads us straight into what I shall call The Researcher's Dilemma, or The Dual Task of the Applied Researcher. This
dilemma can be summed up as follows:
On the one hand, the need to meet external demands on research; to do work and produce materials or reports that are
'immediately useful'that can be incorporated straight into teacher training programmes or into the second language
classroom.
On the other hand, the need to meet the internal demands of the scientific community; to do work that focusses more
directly on theory: generating, building or even refuting existing hypotheses.
In the academic world, there are still those who confuse the distinction between 'basic' and 'applied' research with that of
'research proper' and 'research applied'. For example, Krashen (1983) states that applied research 'attempts to answer practical
problems directly, without recourse to theory'. This is a gross misrepresentation of the nature of applied linguistics. The
requirement that applied research must be theoretically grounded in its choice and formulation of research issues is at least as
categorical as for traditional academic research, writes Nafstad (1982), and continues:
This can plainly be seen in the consequences of a weak or insufficient theoretical basis. A weak or inadequate theoretical
basis in traditional academic research will, at worst, have undesired effects on one's personal academic career, while lack
of theoretical knowledge on the part of the applied researcher may have disastrous effects on the social groups the
particular research is designed to assist.
Particularly for those whose research interests stem from practical work (e.g. teaching), it can be tempting to throw oneself
wholeheartedly into the 'practical' tasks (production of materials, etc). The particular kind of competence needed for such work
is often underratedas is the amount of time it consumes.

< previous page page_214 next page >


< previous page page_215 next page >
Page 215
However, in this area also, it is vital that one accounts properly for the underlying research process with all its theoretical and
methodological considerations. Correspondingly, in work that is more obviously theoretical in nature, care should be taken to
demonstrate if and how the work relates to specific, practical problems. This should be done with respect to the choice and
formulation of research problems, as well as with respect to findings and results.
The research process of applied research is identical to that of basic researchthe differences are found first in the formulation
and justification of research problems (the applied researcher must establish a dual, practical and theoretical rationale for her/his
work), and, secondly, in the reporting phase. The very history of applied linguistics demonstrates the inter-relationship between
theory and practice.
The development in our field, over the last 30 or 40 years, has been one of increasing complexityfrom contrastive analysis,
through the empirical error analyses, and interlanguage analyses, and into what now emerges as a possible new paradigm of
language acquisition research: that which may be called the 'global analysis' (Andenæs, 1984).
In contrastive analysis the potential research areas seemed fairly clear-cut and straightforward: a relatively simple, behaviourist
theory of language learning was paired with a structuralist theory of language, pointing out what should be studied, and how.
The theory having been worked out in advance, the researcher's job was, 'simply to apply it'. Andin theory(!)this work would
result in contrastive analyses of the languages in question, supposedly immediately useful to teachers (and, by implication) to
learners.
We have come a long way since then. Because, as we now know only too well, the theory was too simpleeven if contrastive
studies were helpful in the classroom, they were not helpful enough, where and if they could be successfully made. Since then,
we have moved through the stages of error analysis and interlanguage analysis, discovering on the way more and more about
some of the processes a learner goes through in approaching and acquiring a second language.
On the way, the area of second language research has considerably expanded. While in the 1950's our research was mainly
restricted to the area of linguistic systems and sub-systems (in a restricted, structuralistic sense), it now comprises vastly
different areas. Studies of cognitive style and learning strategies; teacher talk and classroom interaction; socioeconomic factors;
the effect of immersion classes and of mother tongue maintenance; of attitudes and cultural stereotypes; and many more. At the
same time, our whole concept of language has changed considerably: while the phoneme, morpheme and syntax studies are still
vital to our understanding of it, so are pragmatics and text linguistics, bringing in such fields of study as discourse analysis,
communication strategies, global structures of texts, power relations between partners in communication, and so on. There seems
to be no end to what is relevant, or to what needs to be studied.
The process has been one of ever widening the field. We now see language

< previous page page_215 next page >


< previous page page_216 next page >
Page 216
acquisition as the product of an extremely complex interaction of a multitude of factors. A theory of language acquisition, if it is
to be satisfactory, must be able to account for all these factors and for the inter-relations between them.
As Gabriele Kasper pointed out in her comment on Dr Chamot's opening lecture (this volume p.11), research on one
phenomenon in language acquisition is not quite satisfactory as long as its relationship to other phenomena, and (preferably) to a
comprehensive, overall theory of language learning has not been accounted for.
If our research really is to be 'useful', and to point out valid solutions to educational questionsthen we need such a theory. In
order to deal with a practical problem, one must first have a thorough understanding of it. Even if, at present, we knew (or
thought that we knew) what was 'the ideal way' or 'the perfect model', because of the immense variety found among minority
populations (with respect to numbers; age; to their needs and aims; linguistic, cultural and educational background, personality
traits, and to a number of other factors, some of which have probably not even been discovered yetand with respect to economic
resources)we should probably never be able to supply the one Model which can successfully be applied to all students.
What the educational system needs is not one model, and what teachers need is not cookbook recipes of 'how to teach the past
tense'. What is needed, I think, is flexibility. We need teachers who know of more than one way of 'doing it'and who can analyse
the situation of particular students in particular situations. I feel this should be the major goal of teacher training programmes.
However 'theoretical' the work, though, researchers should never forget that second language research is, essentially, applied
research. And while I do not take this to mean that we 'attempt to answer practical problems directly, without recourse to
theory', I do believe that in order to test a theory, we must put it into practice. Also, I think the theory (or the work leading up to
it) should be derived from practice. In my experience, direct co-operation with practising teachersand other users of
researchmakes research less easy, but more validand may provide extremely important input and feedback to 'purely theoretical'
work.
We also need teachers/users to follow our workreceptively and/or critically. And this, then, puts certain obligations on the
researcher, in that she will have to account for her work in such a manner that users can read and understand it. Even if second
language teachers, in my experience, are prepared to devote a great amount of time and energy to developing their professional
competence, they cannot be expected to take the time to study all the details of methodology, etc. which must necessarily be
present in a research report. Also, we cannot write these reportsthat would seriously delay further development of the field.
This aspect of the researcher's dilemma may seem a conflict of loyalties. Loyalty to users (whether they be teachers, students, or
others) indicates focussing on results, findings, conclusions and practical implications. On the other hand, loyalty to the
scientific community dictates that theoretical and

< previous page page_216 next page >


< previous page page_217 next page >
Page 217
methodological problems and critical reservations be emphasised and elaborated. This may well seem 'a waste of time' to those
whose main interests lie in the classroom, and whose main loyalty rests with the users. But this would be a mistake as serious as
that of forgetting the practical relevance of research. If we want the development of our field to continue, both aspects of our
work must be given consideration. It is up to us, then, to turn 'a conflict of loyalties' into 'a dual loyalty'. In order to do this, we
need constant contact and cooperation between users and researchers. Conferences like this one seem to give a great many
opportunities for establishing that.

References
Andenæs, E. (1984) Morsmål, mellomspråk, målspråkhvordan lærer vi nye språk? In A. Hvenekilde and A. Ryen (eds), Kan jeg
få ordene dine, lærer? Oslo: Cappelen.
Krashen, S. (1983) Second language acquisition theory and the preparation of teachers: toward a rationale. Georgetown
University Round Table on Languages and Linguistics 1983: Applied Linguistics and the preparation of Second language
teachers: Towards a rationale, edited by J. E. Alatis, H. H. Stern & P. Strevens.
Nafstad, H. E. (1982) Applied versus basic social research: a question of amplified complexity. Acta Sociologica, 25, No. 3,
259-67.

Kenneth Hyltenstam: Where do we go from here?


The five Nordic conferences on bilingualism that have taken place so far have all, of course, been concerned with a wide range
of research activities in bilingualism, but also with practical aspects of bilingualism in society, perhaps with an emphasis on
educational problems. Since research questions in the area of bilingualism in many cases are so obviously generated by practical
needs in society, I think this linking of theory and practice in conferences such as these is quite natural, necessary and a good
thing.
The flow of information between researchers and practitioners is not, however, always unproblematic. This is in actual fact, to
my mind at least, one of the greater problems in our field. Simultaneously with the accumulation of increasing bodies of
knowledge, people who are practically concerned by these results do not always find their way to them. It is this information
gap and the bridging of it I should like to address in this short note. In other words: where do we go from here if we aim at a
more efficient practical utilisation of the knowledge that exists in our field?
When it comes to application of knowledge gained in research, I shall approach the question against the background of the
Swedish scene, since this is what I am most familiar with and since it is here I can most easily see where improvements might
be possible. I do think, however, that the Swedish scene is not unique in most respects but shows many of the same problems
that can be seen across all the Nordic countries and in Western society in general.
If we first look at the scientific world of bilingualism at an international

< previous page page_217 next page >


< previous page page_218 next page >
Page 218
level, we note that the field as such is extremely broad and comprises a number of branches and specialities and draws on a
number of various disciplines such as Gabriele Kasper just mentioned, or should draw on a number of various disciplines.
Among these branches and specialities, bilingualism and education, the theme of this conference, is one of the most flourishing
ones. The field of second-language acquisition research, which is partly overlapping with bilingualism and education and which
has attracted much attention at this conference, is another extremely active branch of bilingualism in terms of research activities
and the flow of publications.
Now, like in all other scientific fields, the questions addressed in bilingual research are not generated by any sort of algorithm. If
not theory-generated, research questions in many cases have their deepest source in informal observations in real life. Suppose
the most common source is a combination of theory and real-life observations. Depending on his or her theoretical stand and
scientific outlook, the researcher will make certain observations on reality and ignore other things. And this is where I start to
come close to the problems of the research side that I want to discuss.
Much work in the field of bilingualism, like sociopsychological linguistics in general, has been based on or motivated by
hypotheses and notions of bilingual behaviour that are quite broad. Some of the examples of this, which we are all familiar with,
are the following: Are both the bilingual languages stored together or separately? Do different bilingual persons exhibit different
patterns in this respect? Do formal and informal language learning involve different and incompatible learning types as in
Krashen's Monitor Model? Do we find second language learners who operate with 'just' surface fluency and. others who also
attain cognitive academic language skills? Are there speakers who have access to a restricted code only and others who can also
use an elaborated code? Are some speakers semilingual? These are all examples we are very familiar with. code only and others
who can also use an elaborated code? Are some speakers in semilingualism? These are all examples we are very familiar with.
In taking up these notions as broad research questions, I do not intend to criticise what is normal in scientific work, namely to
start the exploration of an area from such broad hypotheses and then, as research proceeds, narrow down the scope to details.
But one problem I encounter here is when these broad hypotheses boil down to common sense beliefs that are made a little
more sophisticated by the theoretical apparatus they are dressed in. This is, of course, not a problem in itself. As I mentioned
before, research questions are to a great extent rooted in informal observations that anyone can make. The problem lies in how
these hypotheses are handled. In all the cases I have mentioned, the hypothesis has been put forward and discussed at length not
only in strictly theoretical publications but also in text books for a broader audience. One cannot blame it only on the reader that
it has not always been clear that the hypotheses are hypotheses and not

< previous page page_218 next page >


< previous page page_219 next page >
Page 219
facts.
In many cases it seems to have been all too tempting for the writer to discuss his or her pet hypothesis as if it were a fact. When
the hypothesis is a formulation of commonly held beliefs, many ordinary people who hear about them will say: 'Yes, this is the
way things must be, this is exactly what I've always believed.' Since the common sense belief now has received status by being
formulated by research, it is much closer to a fact than it was before.
Here I come to the other side of the problem. Why does the ordinary reader take these hypotheses often discussed as facts so
seriously? Why isn't the question more often: 'What is the evidence?' Well, I think as a receiver, when you get the information
that to some extent seems attractive to you, it takes a great deal of critical schooling and critical thinking to ask that question. It
also requires broad knowledge of the field of bilingualism to see through the often convincing argumentation. People who are
practically concerned with bilingualism, such as the whole cadre of teachers, in particular second language teachers and teachers
in bilingual programmes at all levels, but also administrators and consultants, in many cases lack the appropriate training for
this.
If we look at the Swedish scene here, and as I said I think it is similar in other countries, the lack of training among teachers
certainly cannot be blamed on the teachers themselves. In Sweden at least, the interest organisations that exist among them
demand, in a couple of their most prominent programmes, that appropriate and highly qualified teacher training should be
developed and that employment qualifications should be authorised. Rather it has been the neglect of authorities that has
prevented the professionalisation of the area.
Perhaps I paint the picture in too dark colours. There are certainly positive sides in the present development. In Sweden, for
example, second-language teaching programmes are being developed and minority first-language teacher programmes are being
revised, but it is difficult to tell whether future programmes will meet the needs.
Another promising aspect is that an increasing number of administrators take interest in and feel the necessity of knowing more
about the subject area they are administering. There are various signs of this in Sweden, but I will not go into it here.
I will come to the conclusion on the background of this discussion and I can make it in very brief terms: as regards research in
the area of bilingualism and the dissemination of results, more caution should be exercised in clearly stating what are
hypotheses, what are guesses with high truth probability and what are facts. Looking back, it should not be difficult to see how
practice has been thrown from one side to the other by what gets popular acceptance in science at the moment. To remedy the
disadvantages of the practical implementation, I see the professionalism the practitioners themselves are demanding as a goal for
which to work.

< previous page page_219 next page >


< previous page page_220 next page >
Page 220
Euan Reid: Where do we go from here?
Where do we go from here is our overall title for the panel and I offered to take the teacher training aspects of this question. I
am going to try to say something about three aspects of teacher training. Firstly something about structure, then something about
content and lastly something about recruits, the people we take into the teaching profession. What I shall say will, of course, be
based most directly on the experiences of the English situation. It will be related to what I said in my paper (this volume, p. 181)
but I hope that it will make a contribution that may also have some relevance to your rather different concerns in the Nordic
countries.
So let me first come to structure. I would like to start by referring to the first visit that I made to Sweden about five years ago. I
was invited as the only non-Swedish and non-Finnish participant to a Swedish-Finnish conference that was really focussed on
teaching Swedish as a second language, and I was given a brief quite similar to the one that I had at the beginning of this
conference. I did not know very much about the situation; I knew even less about the politics of that particular conference, but I
soon realised when I got there that what was going on was a kind of battle to set up separate structures for training teachers of
Swedish as a second language.
I came in and I realised that I was in the middle of a very hot dispute about the structural changes that many people, whose
work I greatly respected, felt were necessary in establishing a greater professionalism in the teaching of Swedish as a second
language. Now you may realise my dilemma, given that I was singing a very similar song to the one that I sang to you on
Tuesday. I was talking about the experience that we have had of setting up separate structures and then of beginning to retreat
from these. We were at a slightly earlier stage, but I realised that what I was going to say was not quite what the people who
had paid my fare there were hoping to hear. However, I said it and they were very polite and quite interested. But I think they
went ahead anyway and set up the separate structures for training programmes for Swedish as a second language.
Now it seems to me there is a great difficulty if we accept the kind of direction of change that I was describing the other day. If
we accept that it is necesssary, perhaps inevitable, for very strong general educational reasons as well as for reasons to do with
the most effective processes of language learning in the school setting in a multilingual population, if we accept that, then I
think we have to look again rather closely at the kind of separate structures that we have in teacher training. Because at the
moment, what we have certainly at secondary level is a situation where you choose in many cases a single specialismEnglish as
a second languagefrom among a series of options like science or English mother tongue, history, etc. So you choose 'English as
a second language'. The problem is, of course, that, as many people are realising, it is not exactly clear what English as a second
language is as distinct from English. As we are increasingly dealing with children who are not beginning from scratch but
certainly from a secondary

< previous page page_220 next page >


< previous page page_221 next page >
Page 221
level with considerable experience already of English-medium education, perhaps also of separate English teaching classes, it
hardly seems appropriate to be continuing with structures which train teachers to teachI'll caricature it a little bit to make the
point clearlyrelatively homogeneous groups of learners in a kind of foreign language teachingpedagogical style.
I think, as I said, I am caricaturing a little bit, but that is what we are still stuck with when we have courses which in the English
situation are now often called 'English for speakers of other languages', that we attempt together to prepare people to teach
English abroad and teach English as a second language in England. Meanwhile, of course, along the corridors of our training
institutions there are other people training to be specialists in teaching English as mother tongue and, perhaps, in the same
corridor a little further along there are people training to be teachers of foreign languages, in our case mostly of French, but still
there are also some other languages. And they are all being trained in rather separate divisions, in rather separate ideologies, I
would say. They have different views of language learning, and they are offered different views of language learning,
appropriate classroom procedures and so on. Of course in the end they are all teaching the same children. The children are
receiving second language instruction from one of the teachers, foreign language from another, perhaps mother tongue from yet
another. I forgot the fourth group who, in our case are not very often there on the corridorthe teachers of minority languages. I
think they are more often here (Scandinavia) in your colleges.
Now the logic of the structural changes that I have been describing in second language teaching in schools is towards an
increasing integration with main stream teaching. And I think it is clear that we have to look to a time when second language
teaching as a separate institution and second language teachers as separate specialists will perhaps even cease to exist. It is a
terrible thought if we are part of this branch of the profession. Some people say already: 'Does ESL exist?' One of my groups of
students who had just finished a year's training actually came up with that question at the end.
In the institution where I work we are already considering a somewhat different approach to the training structure. One option
we are considering is the preparation of language educators with a common corewhich I might come to in my second section in
a minuteof understanding and preparation to do with language learning in the school settings and then taking various
combinations of language subjects so that people become second langauge and mother tongue specialists or second language,
let's say, for example, English as a second language and English mother tongue, or English as a second language and a minority
language, or English as a second language and a foreign language. So what we have is a common core and then people would
choose some particular sub-specialisms within that. That is one model that I think is a very fruitful direction for future
experiments, perhaps indeed it is one that you have already in place.

< previous page page_221 next page >


< previous page page_222 next page >
Page 222
Now my second sub-heading was content, related to the role of the second language teacher. In fact I am going to be fairly brief
with this since several of the previous speakers have already touched on it. It is one advantage of coming late in a panel like
this. One pressure towards taking English as a second-language teaching preparation out of initial training altogether is the
changing role of such teachers, if we accept again the model that I have described of integration in schools. If the work is
moving much more towards a role for the teacher which is collaborative, co-operative, advisory, involved in negotiating with
science teachers, history teachers and so on, then some people feel that this is really not an appropriate role at all for the young
probationary teacher just finishing initial training, lacking perhaps in professional maturity, certainly lacking in the kind of status
needed to negotiate on equal terms with colleagues in the secondary schools, who are themselves experienced teachers. And
another option we are considering is actually recommending that the ESL teacher specialism be taken out of initial training
altogether and left entirely to in-service teacher training, so that people would have simply a general training in one of the
mainstream subjects, curriculum subjects, and then would come back in and add whatever second language expertise is needed
to the original training. Now, what is the nature of this new expertise? This, I think, is what has already been covered, so I will
say very little about it. It is clear that we are no longer talking simply about teaching people about the grammar and the
phonology of the language they are going to teach and something about language teaching methodology. We have to extend into
the kind of things which Gabriele Kasper was talking about at the beginning, to do with communication break-down and repair,
with inter-ethnic communication and with things like racism awareness, and things like curriculum change strategies; they seem
to me new areas which I have to throw into this already rather large teacher-training agenda.
But let us come to my last point: recruitment for teacher training. This is a very obvious one. If we have that kind of view of the
content and structure of teacher training, it seems to me that probably the best people to train for this kind of language educator
are those who are themselves actively bilingual. Now that, of course, will include manyit need not be exclusivelybut it really
ought to include many more people than we have at the moment who are speakers of languages of the relevant minority
communities. I have done for the last few years a survey at my institution, and our general intake into secondary teacher training
is about 500 each year. When you get down to it, there are only about 15 to 20 of these who are speakers of the relevant
languages. Now that is a very shameful confession, but I think again it is not untypical of institutions in other parts of England
and perhaps in some of your countries too. It seems to me it becomes increasingly hard to defend a section of our profession,
composed as we are at the moment predominantly of people like me and many of you, that is to say people who are native
speakers of the host language and who

< previous page page_222 next page >


< previous page page_223 next page >
Page 223
are interested in language, know quite a lot about language and language learning, but who are not speakers of the other
languages too.
So to sum up on where we go from here, let me say where I hope we might be in five years. A teacher training college offers an
integrated preparation for language educators rather than a series of separate streams for the different sub-specialisms. Such
recruits will be offered a common core, drawn not only from second language acquisition and, I would say, sociolinguistic as
well as psycholinguistic sources, but would also have quite a strong emphasis on things to do with classroom dynamics, things to
do with school dynamics, negotiating skills, and so on. And lastly we shall be recruiting a rather high proportion of the entrants
from minority students themselves.

Anna Uhl Chamot: Where do we go from here?


I have listened with a great deal of interest not only to the contributions of the panel this morning with some very important
ideas which I think we shall all take seriously, but also to the presentations that have taken place during the week and most
especially to the informal conversations that I had the privilege of having with many of you. I feel I have been very fortunate to
have been here and to have got some kind of a global understanding of many of the issues that are of concern in your countries
andstrangely enoughin my country as well. There seem to be so many aspects in common. I am thirsty for more information. I
think when a group like this comes together we share ideas, we discuss some of the problems we have encountered, we talk
about possible solutions, we share some solutions that some of us may have to problems that others are encountering for the
first time. We all seem to want to have more information about things like looking at the programmes that we have for our
linguistic minorities in the various countries, and really evaluating them, following up those children on a longitudinal basis,
finding out what really happens to them in their subsequent schooling after they are out of the special programme and in life, in
fact. How do we know that the programmes that we are supporting really are effective, and how can we find out if they are?
Another issue that I want to know more aboutand I think that many of us dois different instructional approaches to developing
competence in the second language. I think, as several people have mentioned, we will pass the point of saying there is one truth
and one answer and one approach that is good for everyone. But how do we compare the approaches, do we know which
approaches in the classroom work best for which kinds of students and which kinds of situations given certain kinds of
educational goals? I think we all need to know more about that. In addition to the actual instructional approaches, what kinds of
affective considerations do we need to take into account as we work with linguistic minority children? Then, a situation that I
have heard ofand this has been informally from a number

< previous page page_223 next page >


< previous page page_224 next page >
Page 224
of peopleand which is certainly a burning question in the United States, is how do we meet the needs of older school children
who are illiterateor as we tend to say in the States: preliterate, because we hope they are going to be literatein their native
language? What is the best approach with these students, and should we teach the literacy skills first in their mother tongue
before we begin literacy in the second language, or if it is an adolescent, is there time to do that, is it worth taking the time, is
there a way to by-pass that and go directly for literacy in the second language? What is the best way of meeting this problem?
And now in the United States we certainly do not know what the best way is. We are experimenting with a number of things,
but to my knowledge we are not doing any research in this area. This is really a shame because it is an area that I think all of us
are going to be encountering more and more as we have more immigrants, more refugees.
Another area: I think we tend to agree that the role of prior knowledge of students is extremely important, that what they already
know is going to guide them into new information about language and a great deal of the content curriculum that we are trying
to teach students. How do we assess what students already know if as in the case when you started teachingyou do not know
their language? How do you find out what they already know in the assessment of what prior knowledge can be built on to
develop their knowledge in their second language? So we need, I think, more information about assessing prior knowledge that
has been attained in the first language. Another area that we need to know much more aboutwhich a number of the people on
the panel this morning and other people throughout the conference have indicatedis what are the cognitive processes involved in
learning language and in learning content matter? How similar are they? Are there certain different kinds of cognitive processes
that relate only to language or are there general cognitive processes that include both language and subject matter? And how can
we find out more about them? And, finally and very, very important: how do cultural differences affect school success? We
have many different kinds of cultures. One of the interesting things for me has been finding out about some of the kinds of
linguistic minorities that you have in your schools, that we have either none or very few of in our schools. As I mentioned on
the first day, our majority is Spanish-speaking students and after that come the speakers of the South East Asian languages. We
are beginning to get some expertise with languages and cultures like Vietnamese, Cambodian, Laotian, and Hmong, and with
Chinese, because we have had a Chinese population in the United States for a very long period. But there are so many other
languages that we have small pockets of, that we do not know really nearly as much as you do about the background and culture
of Turkish immigrants or Iranian immigrants or immigrants from other areas that we have relatively few of at this point in the
United States. Do different cultural groups succeed more or less well in school, depending on the cultural distance they
encounter, or the social

< previous page page_224 next page >


< previous page page_225 next page >
Page 225
distance in Schumann's terms that they perceive between themselves and the host country and the culture of the host country?
Logically we would think that the less the social-cultural difference, the easier it would be for students to be successful in the
new school system, and the greater the social-cultural distance, the more difficult it would be.
So all these questions are really a need for more information. What would I like to propose is a bit practicalI hope it is practical.
We have been talking about practical goals, collaborative research. I think the general theme of what we have all been saying
this morning is the need for cooperation, whether it is between disciplines, between researchers and teachers, between different
kinds of teachers who are being trained in areas of dealing with language issues, the thing seems to be co-operation.
I think the next step for co-operation is that we all need to continue to have a way of disseminating information, exchanging
information. We all do this on an informal basis, we make contacts with other professionals at conferences like this, we all have
our wonderful yellow sheet with everyone's name and address which we can all make use of when we go home, and we have
written down people's addresses who want certain information from us and we promised to send them things and we will keep
those promises. That is fine: that is an informal network and it is very important. What I would like to propose is that I think we
really need a more formal kind of information network. So much is going on in every country that we have heard about. We all
seem to have similar needs, similar problems. Some wonderful things are being done in every country, but we just know a little
bit about them because we have been here this week. So I would like to propose a mechanism for an information exchange on
an international basis. And those of you who are familiar with the National Clearinghouse for Bilingual Education in the United
States will recognise the model that I am proposing.
I think it needs to go beyond the national level, however, and become international, something that will be called something like
'The international centre for information on the education of linguistic minorities'; in other words, whatever kind of education,
whether it is second language education, whether it is bilingual education, whatever is done for linguistic minorities in different
countries. The kinds of information that will be collected by such a centre will be things like information about models of
programmes. This morning's plenary was extremely interesting and different models were discussed. I, for one, would like to
have much more information about those models. Also one of the kinds of information that could be collected by such an
information centre would be descriptions of programme models and secondly descriptions of instructional methods. We are all
constantly adapting and revising methods and seeing how they work with certain kinds of students and certain kinds of
situations. Those kinds of description are important to disseminate.
Third: the results of all the wonderful research that is going on now and

< previous page page_225 next page >


< previous page page_226 next page >
Page 226
is projected for the future. To answer the question'what is the evidence?' Again to stress what you were saying about more
cautionwe need it. Yes, we need to know about research, but we need the research to be reported in a form that is useful for the
practitioner. In other words we need to digest practitioner-oriented research summaries, accurately done by people who
understand research and who will take the necessary caution not to report as fact what is really hypothesis. But to provide this
kind of information to practitioners through research-summaries I think is extremely important.
The fourth kind of information that can be collected and disseminated by such a centre would be information about instruction
materials, especially materials in the less-commonly taught languages, for the countries and the areas that do have mother
tongue instruction. Those of you who have been involved in mother tongue instruction know how difficult it is to acquire
materials. I mean it is easy in the major languages of the world to acquire instructional materials, but when you get into the
uncommonly taught languages, e.g. Gujarati or Turkish or Farsi, what kinds of instructional materials in those languages are
appropriate for linguistic minority children living in a new environment, no longer in their native country? For instance, in the
U.S.A. we have lots of information about materials for our major groups, but we certainly do not have any information at all
about materials for some other groups of which we have just a small number. We would love to know of some of the materials
that countries that have large populations of those groups are using, to see if they would be useful for our needs. And it would
be wonderful to have a centre that we could write to and say: Please, how can we order copies of these books to see if they will
work with our children?
Fifth: information about teacher training. As you mentioned, information about teacher training, both for language specialist
teachers, generic language specialists or teacher training models and also, I think increasingly, what kinds of in-service or
perhaps even pre-service training for the mainstream regular classroom teacher who is increasingly going to be receiving more
and more linguistic minority children. What kinds of training options are available for those teachers as well?
And the sixth thing simply: Information about the different cultural groups. Suddenly a new cultural group comes to your
country. Where would you turn immediately for information about their value systems, their customs, their background, their
educational system. All this kind of information which is essential in planning for their needs.
So I think that I would like to propose that we put our minds together and see how we could fund such a centre that would
collect, analyse and disseminate information about the education of linguistic minorities.

< previous page page_226 next page >


< previous page page_227 next page >
Page 227
Gabriele Kasper: Additional Comment
There is a saying that goes: Beware when all men/women agree with you! I was most surprised how much in accordance the
opinions that were expressed on this panel really are. One of the things that I would underline has to do with the various remarks
that have been made about teacher training. It seems to me worth underlining indeed that the main goal in teacher training has to
be making teachers aware of the teaching situation, their own teaching situation, their own assumptions about teaching, their
own implicit theories about what bilingual education is or should be like in order for them, on that basis, to assess whether their
assumptions and their teaching procedures, their didactic and methodological decisions are the appropriate ones. This, I think,
must be the goal of all teacher training, enabling teachers autonomously to make such decisions. And our task as researchers and
teacher trainers should be to provide the teachers with such an ability as well as we can, rather thanas has been suggestedto
provide teachers with recipes which will not work anyway.

Ellen Andenæs: Additional Comment


I just want to add one point, in relation to some questions that have been asked here: 'Does ESL exist anymore?', 'Is second
language teaching going to die?' If that were true, we could just stop this conference and go home, couldn't we? I think that
whatever models we come up with (for education in general or language teaching in particular), there will always be a need for
second language teacherssimply because there will always be a need for teachers who have been trained to view the target
language in this particular perspective, and to analyse the situation of the student.

< previous page page_227 next page >


< previous page page_ii next page >
< previous page page_229 next page >
Page 229

Appendix
This volume is one of three which comprise the Proceedings of the Fifth Nordic Conference on Bilingualism, held in
Copenhagen, Denmark, 22-25 June 1987. The other two volumes appear in book form, their titles and contents being as follows:
HOLMEN, ANNE, HANSEN, ELIZABETH, GIMBEL, JØRGEN AND JORGENSEN, J. NORMANN (EDS) (1988)
BILINGUALISM AND THE INDIVIDUAL. COPENHAGEN STUDIES IN BILINGUALISM, VOL. 4. CLEVEDON:
MULTILINGUAL MATTERS.

Contents
Anne Holmen, Elisabeth Hansen, Jørgen Gimbel and J. Normann Jørgensen: Introductory Note
Interactimal Analysis
Karen Margrethe Pedersen: Second Language Learners in the German Minority in Denmark
Hans Dahlbäck: Children's Questions to Children and Adults in a Second Language
Tuula Hirvonen: Monolingual and Bilingual Children's Foreigner Talk Conversations
Jehannes Ytsma: Bilingual Classroom Interaction in Friesland
J. Normann Jørgensen: Turkish Children's Communication Strategies in Danish
Elizabeth Lanza: Language Strategies in the Home: Linguistic Input and Infant Bilingualism
Hans Vejleskov: A Critique of the Distinction Between 'Academic' and 'Communicative' Language,from a Pragmatic Point of
View
Lenore Arnberg: Assessing Comprehension Skills in Preschool Bilingual Children
Gisela Håkansson and Inger Lindberg: What's the Question? Investigating Questions in Second Language Classrooms
Maija Kalin: Metalinguistic Knowledge and Understanding in Adult Language Learning

< previous page page_229 next page >


< previous page page_230 next page >
Page 230
Linguistic Structure and Variation in Bilinguals
Barbara J. Boseker: Bidialectalism in the United States
Gunnel Melchers: Bidialectism and the School with Special Reference to the Shetland Experience
Miodrag Stankovski: Obstruents and their Development in the Population D07. In the Archive for Diaspora Languages
Ulla Laurén: Tvåspråkiga Barns OrdförrådInterferens och Integration
Solveig Strömman: Fackslang i Tvåspråkiga FöretagSvensk eller Finsk?
Niels Haastrup: Spelling Errors in Danish Loan Words or on the Changing Linguistic Dominance and Consciousness of
Language History
Different Backgrounds and 2nd Language Learning
Edith Mägiste: Leaning to the Right: Hemispheric Involvement in Two Immigrant Groups
Christopher Stroud: Literacy in a Second Language: A Study of Text Construction in Near-Native Speakers of Swedish
Raymond Mougeon and Edouard Beniak: Minority Language Schooling without Home Language Maintenance: Impact on
Language Proficiency
Anne Hvenekilde: Elevene fra Språklige Minoriteter og Matematikkfaget
Peter Berliner: Cognitive Style and Attitudes in Bilingual Inuits
Bent Søndergaard: Motivationsprofilen ved Indlæring af L2 med henblik på Dansk på Færøerne
Literature and Bilingualism
Marie-Alice Séférian: Kan en Digter være Tosproget? Tilfældet Kateb Yacine
Harald Gaski: Joik, Etno-Poesi og Majoritetens Forståelse
Kjell Herberts: Språkmöten i Minoritetslitteraturen
Inge Kleivan: Børnebøger i Grønland: Sprog og Indhold

< previous page page_230 next page >


< previous page page_231 next page >
Page 231
JØRGENSEN, J. NORMANN, HANSEN, ELIZABETH, HOLMEN, ANNE AND GIMBEL, JØRGEN (EDS) (1988),
BILINGUALISM IN SOCIETY AND SCHOOL: COPENHAGEN STUDIES IN BILINGUALISM, VOL. 5. CLEVEDON:
MULTINGUAL MATTERS.

Contents
J. Normann Jørgensen, Elisabeth Hansen, Anne Holmen and Jørgen Gimbel: Introductory Note
Bilingualism in Society and School
Emilia Nercissians and Caro Lucas: A Game Theoretical Approach to Language Planning
Christer Laurén: Societal Bilingualism and Language Planning
Lenora A. Timm: Against All Odds: Language Promotion Efforts in Brittany
Charlotte Hoffmann: Linguistic Normalisation in Catalonia: Catalan for the Catalans or Catalan for Catalonia?
Doris Pyee-Cohen: On the Planification of Language Acquisition
Status of Languages
Emilia Nercissians: Bilingualism with Diglossia: Status and Solidarity Dimensions
Jeannine Gerbault: Language Use and Attitudes: The Rise of Sango
Malan Saleem: Tosprogethed på Færøernei historisk perspektiv
Jeffrei Henriksen: Faktorer, der fremmer eller truer det færöske sprog
László Szabó: Malecite Bilingualism
Language Ecology and the School
Aquigssiaq Møller: Sprogpolitik og sprogplan-lægning efter hjemmestyrets indførelse
Grete Ulrich: Børn, teenagers og de to sprog i Grønland
Viking Brunell: Tvåspråkighet och skola i Svenskfinland
John de Vries: Language Use by French Immersion Graduates
Raymond Mougeon and Édouard Beniak: Should the French-Canadian Minorities Open Their Schools to the Children of the
Anglophone Majority?
Ellen Andenæs: Bilingual Education or Education for Bilingualism? Some Notes from Norway
Roger Källström: Bilingual Education and Bilingualism in the Swedish Comprehensive School

< previous page page_231 next page >


< previous page page_232 next page >
Page 232
Group Bilingualism
Mehroo Northover: Bilinguals and Linguistic Identities
Dorrit Poulsen Hawkesworth: Incongruity of Sexual Norms and Behaviour in the Danish Schools: Notes for Discussion
Jesper Hermann: Bilingualism Versus Identity
Gunnel Knubb-Manninen: Andraspråksinlärning i olika subkulturer
Pre-school Children
Ewa A. Chylinski: Flygtningeførskolebørnmål og midler i tosproglig opdragelse
Ole Kragh: Tokulturelt forsøgsprojekt i børnehaven Svalen
Søren C. Olesen: Om de 0-6 årige flygtningebørns personlighedsudvikling og integration
Britt-Ingrid Stockfelt-Hoatson: Olika metoder att förbättra förskolan för barn med annat hemspråk än svenska
Classroom and Practice
Lis Glebe-Møller: Om hensynet til voksne indvandreres kulturelle og uddannelsesmæssige forudsætninger i undervisningen
Heri Kragesteen and Karen Skovbjerg: Overvejelser over introduktion af andetsprog i den færøske skole
Per Pinholt and Leif Sand: Indvandrerundervisningens behov
Mijo Tomasevic * Lingvistiska aspekter på hemspråksundervisningen i den svenska skolen
Ulla Varming: Workshop om tosproget undervisning

< previous page page_232 next page >


< previous page page_233 next page >
Page 233

Index

A
Adverbs 97-113
B
Bilingual Education 11-35
-early 193-202
-typology of 169-176
Bilingual teachers 181-191
Bilingualism 141-149, 203-210

C
Catalonia 193-202
Competence 141-149
Conversation 61-66
Creolisation 129-140

D
Danish 85-96, 177-179
Denmark 151-167

E
Education 203-210
ESL 43-60
-content-based 43-60
Ethnic identity 203-210

F
Fossilisation 67-84
France 193-202
Functional approach 85-96

G
German as a second language 141-149
Greenland 177-179

I
Immaturity 115-128
Immigrant children's language learning 129-140
Immigrant varieties 129-140
Information structure 85-96
Input 97-113
Intellectual language 61-66
Intercultural Education 141-149
Interference 115-128
Introspective methods 37-42
Italy 193-202

L
Language
-change 129-140
-planning 177-179
-policy 177-179
-shift 203-210
-varieties 141-149
Learning strategies 11-35, 37-42, 43-60
LEP students 11-35
Lexis 67-84
Longitudinal study 85-96

M
Mainstreaming 181-191
Minority Education 169-176
Mothertongue maintenance 151-167
Motivation 115-128

N
National minorities 169-176
Negation 97-113

P
Placement rules 97-113
Pragmatics 61-66
Prefabricated chunks 97-113
Prefabricated patterns 115-128
Pre-school children 61-66
Procedural knowledge 37-42

R
Research methods 37-42

S
Second language acquisition 85-96
-adult 97-113
Second language proficiency 67-84
Social language 61-66
Sources of difficulty 115-128
Spain 193-202
Strategic competence 37-42
Swedish 129-140
Syntax 85-96

T
Teacher training 11-35, 177-179, 181-191

U
U.S.A. 11-35
Utterance function 61-66

< previous page page_233 next page >


< previous page page_234
Page 234
V
Verbal reports 37-42

Y
Yugoslav migrant children 151-167
Yugoslavia 169-176

Z
Zagreb Project 115-128

< previous page page_234

Vous aimerez peut-être aussi