Language, ethnicity & Education: Case Studies On Immigrant Minority Groups and Immigrant Minority Languages (Multilingual Matters)

Citation preview

cover

title: author: publisher: isbn10 | asin: print isbn13: ebook isbn13: language: subject publication date: lcc: ddc: subject:

next page >

Language, Ethnicity, and Education : Case Studies On Immigrant Minority Groups and Immigrant Minority Languages Multilingual Matters (Series) ; 111 Broeder, Peter.; Extra, Guus. Multilingual Matters 185359430X 9781853594304 9780585126067 English Linguistics minorities, Sociolinguistics, Language and education, Linguistic demography. 1999 P40.5.L56B76 1999eb 408/.6/93 Linguistics minorities, Sociolinguistics, Language and education, Linguistic demography.

cover

next page >

< previous page

page_i

next page > Page i

Language, Ethnicity and Education

< previous page

page_i

next page >

< previous page

page_ii

next page > Page ii

MULTILINGUAL MATTERS SERIES Series Editor Professor John Edwards, St. Francis Xavier University, Antigonish, Nova Scotia, Canada Other Books in the Series Beyond Bilingualism: Multilingualism and Multilingual Education JASONE CENOZ and FRED GENESEE (eds) Language Attitudes in Sub-Saharan Africa EFUROSIBINA ADEGBIJA Language Planning: From Practice to Theory ROBERT B. KAPLAN and RICHARD, B. BALDAUF Jr. Language Reclamation HUBISI NWENMELY Linguistic Minorities in Central and Eastern Europe CHRISTINA BRATT PAULSTON and DONALD PECKHAM (eds) Quebec's Aboriginal Languages JACQUES MAURAIS (ed.) The Step-Tongue: Children's English in Singapore ANTHEA FRASER GUPTA Three Generations - Two Languages - One Family LI WEI Other Books of Interest Adult ESOL Learners in Britain A.L. KHANNA, M.K. VERMA, R.K. AGNIHOTRI and S.K. SINHA Encyclopedia of Bilingual Education and Bilingualism COLIN BAKER and SYLVIA PRYS JONES Ethnicity in Eastern Europe SUE WRIGHT (ed.) Language, Culture and Communication in Contemporary Europe CHARLOTTE HOFFMANN (ed.) Managing Language Diversity S. WRIGHT and KELLY-HOLMES, H. (eds) Monolingualism and Bilingualism SUE WRIGHT (ed.) Please contact us for the latest book information: Multilingual Matters, Frankfurt Lodge, Clevedon Hall, Victoria Road, Clevedon, BS21 7HH, England " target="_BLANK">http://www.multilingual-matters.com

< previous page

page_ii

next page >

< previous page

page_iii

next page > Page iii

MULTILINGUAL MATTERS 111 Series Editor: John Edwards

Language, Ethnicity and Education Case Studies on Immigrant Minority Groups and Immigrant Minority Languages Peter Broeder and Guus Extra MULTILINGUAL MATTERS LTD Clevedon • Philadelphia • Toronto • Sydney • Johannesburg

< previous page

page_iii

next page >

< previous page

page_iv

next page > Page iv

Disclaimer: This book contains characters with diacritics. When the characters can be represented using the ISO 8859-1 character set (>" target="_BLANK">" target="_BLANK">http://www.w3.org/TR/images/latin1.gif">http://www.w3.org/TR/images/latin1.gif), netLibrary will represent them as they appear in the original text, and most computers will be able to show the full characters correctly. In order to keep the text searchable and readable on most computers, characters with diacritics that are not part of the ISO 8859-1 list will be represented without their diacritical marks. British Library Cataloguing in Publication Data A CIP catalogue record for this book is available from the British Library. ISBN 1-85359-430-X (hbk) Multilingual Matters Ltd UK: Frankfurt Lodge, Clevedon Hall, Victoria Road, Clevedon BS21 7HH. USA: 325 Chestnut Street, Philadelphia, PA 19106, USA. Canada: OISE, 712 Gordon Baker Road, Toronto, Ontario, Canada M2H 3R7. Australia: P.O. Box 586, Artamon, NSW, Australia. South Africa: PO Box 1080, Northcliffe 2115, Johannesburg, South Africa. Copyright © 1999 Peter Broeder and Guus Extra 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. Printed and bound in Great Britain by Short Run Press Ltd.

< previous page

page_iv

next page >

< previous page

next page >

page_v

Page v

Contents Preface

vii

Introduction 1 Demographic perspectives: Case studies on immigrant minority groups in population statistics

1 1

1.1 Aims and method 1 1.2 European Union statistics on immigrant minority groups 2 1.2.1 Comparative European Union statistics 4 1.2.2 The Netherlands as case study 8 1.3 Census experiences in English-dominant immigration countries 8 1.3.1 Australia 11 1.3.2 Canada 13 1.3.3 The United States 16 1.3.4 South Africa 19 1.4 Conclusions and discussion 2 Sociolinguistic perspectives: Case studies on the vitality of immigrant minority languages

23 23

2.1 Aims and method 24 2.2 Language-related criteria as determinants of ethnicity 24 2.2.1 Feasibility study 29 2.2.2 Validation study 34 2.2.3 Evaluation 35 2.3 Home language survey 40 2.4 Status and use of the most frequently mentioned languages 41 2.4.1 Turkish

43 2.4.2 Arabic 47 2.4.3 Berber 48 2.4.4 Papiamentu 51 2.4.5 English 53 2.4.6 Sranan/Surinamese 57 2.4.7 Chinese 59 2.4.8 Malay 61 2.4.9 Spanish 63 2.4.10 Hind(ustan)i 67 2.4.11 Romanes 68 2.5 Conclusions and discussion

< previous page

page_v

next page >

< previous page

next page >

page_vi

Page vi 3 Educational perspectives: Case studies on the status of immigrant minority languages in education

73 73

3.1 Aims and method 74 3.2 Educational experiences in European Union countries 74 3.2.1 Belgium 77 3.2.2 France 82 3.2.3 Germany 87 3.2.4 Great Britain 93 3.2.5 The Netherlands 100 3.2.6 Sweden 105 3.3 Conclusions and discussion References

111

Appendices 1 Home language survey: instruction and questionnaire

121

2 Questionnaire on IMLI in compulsory education

125

3 Directive 77/486 of the CEC (1977)

129

< previous page

page_vi

next page >

< previous page

page_vii

next page > Page vii

Preface As a consequence of processes of migration and minorization, the trend in many European countries is towards increasing cultural and linguistic diversification. These processes of cross-national change and their cross-continental equivalents in the past are the focus of this study. For a number of years, we have addressed the theme of language, ethnicity, and education in a European context of immigrant minority groups and immigrant minority languages. During these years, we have taken an interdisciplinary approach addressing this theme from demographic, sociolinguistic, and educational perspectives. In doing so, we were greatly inspired by such colleagues as Michael Clyne in Australia, Joshua Fishman in the USA, John de Vries in Canada, and Rajend Mesthrie in South Africa, who devoted much of their lives in different parts of the world to studying the theme of language, ethnicity, and education in English-dominant environments. In this volume, we approach this theme not only from different disciplinary perspectives, but also from different scale levels and different sources of evidence. A common characteristic of the perspectives chosen is the case study approach. A demograpic perspective is taken in Chapter 1. It includes case studies on immigrant minority groups in population statistics of both European and non-European immigrant countries. The main issue in this chapter is the role of language and ethnicity in multicultural population statistics. A sociolinguistic perspective is taken in Chapter 2. Case studies on the vitality of a variety of immigrant minority languages are presented, which were derived from a relatively large-scale home language survey of immigrant minority children in one European Union country, i.e., the Netherlands. The chapter gives an overview of language-related criteria as determinants of ethnicity, the design and outcomes of the home language survey, and the status and use of the most frequently mentioned community languages. An educational perspective is taken in Chapter 3, which presents case studies from different European countries on the status of immigrant minority languages in education. For each country, the focus is on arguments and objectives, target groups and enrolment, and implementation and organization of immigrant minority language instruction (referred to as IMLI). The evidence presented in this volume comes from different sources. Chapters 1 and 3 are largely based on secondary analyses of available data and references, complemented in Chapter 3 by oral or written information supplemented by key informants. Chapter 2 is based on the collection and analysis of primary empirical data. The aim of all the chapters is to present a detailed cross-national and cross-linguistic documentation of the available evidence. Apart from the informants who were most helpful in collecting the evidence presented in Chapter 2 (children, teachers, and principals of many primary schools in the Netherlands) and Chapter 3 (colleagues in different European Union countries), we want to acknowledge Roeland Van Hout for his methodological input, and Gwen Perret and the Language Center at Tilburg University for their editorial

< previous page

page_vii

next page >

< previous page

page_viii

next page > Page viii

support. The accomplishment and publication of this study would not have been possible without the financial support of the European Cultural Foundation in Amsterdam, the Dutch Ministry of Education, Culture and Science, and the Provincial Board of Northern Brabant, the Netherlands. PETER BROEDER AND GUUS EXTRA RESEARCH GROUP ON LANGUAGE AND MINORITIES TILBURG UNIVERSITY, THE NETHERLANDS

< previous page

page_viii

next page >

< previous page

page_1

next page > Page 1

1 Demographic perspectives: Case studies on immigrant minority groups in population statistics 1.1 Aims and Method The focus of this chapter is on societies that have become or are becoming multicultural and multilingual as a consequence of processes of immigration and minorization. From a demographic point of view, four types of criteria have been proposed and used for identifying the multicultural composition of such societies in terms of different ethnocultural population groups, i.e., nationality, birth-country, self-categorization, and home language use. In Section 1.2., we will deal with European Union statistics on immigrant population groups. As yet, European-wide statistics have only been made available in terms of nationality, although comparative cross-national birth-country data are being prepared. We will focus on available comparative EU statistics and on the Netherlands as a case study. Given the decreasing intergenerational significance of nationality and birth-country criteria for the multicultural composition of immigrant societies, complementary or alternative criteria have been used in various countries with a longer immigration history. In Section 1.3, we will focus on four non-European English-dominant immigration countries with a relatively long history of collecting nation-wide demographic data, derived from written questionnaries, in periodical censuses. We will deal in particular with the identification of ethnocultural minority groups by means of census questions about ethnicity and language use. Conclusions and a discussion of the data will be presented in Section 1.4. 1.2 European Union Statistics on Immigrant Minority Groups As a consequence of socio-economically or politically determined processes of migration, the traditional patterns of language variation across Western Europe have changed considerably over the past several decades (cf. Extra and Verhoeven 1998). The first pattern of migration started in the sixties and early seventies, and it was mainly economically motivated. In the case of Mediterranean groups, migration initially involved contract workers who expected - and were expected to stay for a limited period of time. As the period of their stay gradually became longer, this pattern of economic migration was followed by a second pattern of social migration as their families joined them. Subsequently, a second generation was born in the immigrant countries, while their parents often remained uncertain or ambivalent about whether to stay or to return to the country of origin. These

< previous page

page_1

next page >

< previous page

page_2

next page > Page 2

demographic shifts over time have also been accompanied by shifts of designation for the groups under consideration - 'migrant workers,' 'immigrant families,' and 'ethnic minorities,' respectively. As a result, many industrialized Western European countries have a growing number of immigrant populations which differ widely, both from a cultural and from a linguistic point of view, from the mainstream indigenous population. In spite of more stringent immigration policies in most European Union (EU) countries, the prognosis is that immigrant populations will continue to grow as a consequence of the increasing number of political refugees, the opening of the internal European borders, and political and economic developments in Central and Eastern Europe and in other regions of the world. It has been estimated that by the year 2000, about one third of the population under the age of 35 in urbanized Western Europe will have an immigration background. 1.2.1 Comparative European Union statistics. Within the various EU countries, four major immigrant groups can be distinguished: people from Mediterranean EU countries, from Mediterranean non-EU countries, from former colonial countries, and political refugees (cf. Extra and Verhoeven 1993a, 1993b). Comparative information on population figures in EU member states can be obtained from the Statistical Office of the EU in Luxembourg (EuroStat). An overall decrease of the indigenous population has been observed in all EU countries over the last decade; at the same time, there has been an increase in the immigration figures. Although free movement of migrants between EU member states is legally permitted and promoted, most immigrants in EU countries originate from non-EU countries. According to EuroStat (1996), in January 1993, the EU had a population of 368 million, 4.8% of whom (almost 18 million people) were not citizens of the country in which they lived. The increase in the non-national population since 1985 is mainly due to an influx of non-EU nationals, whose numbers rose from 9 to 12 million between 1985 and 1992. The largest numbers of immigrants have been observed in France, Germany, and Great Britain. For various reasons, however, reliable demographic information on immigrant groups in EU countries is difficult to obtain. For some groups or countries, no updated information is available or no such data have ever been collected at all. Moreover, official statistics only reflect immigrant groups with legal resident status. Another source of disparity is the different data collection systems being used, ranging from nation-wide census data to more or less representative surveys. Most importantly, however, the most widely used criteria for immigrant status nationality and/or country of birth - have become less valid over time because of an increasing trend toward naturalization and births within the countries of residence. In addition, most residents from former colonies already have the nationality of their country of immigration.

< previous page

page_2

next page >

< previous page

page_3

next page > Page 3

There are large differences among EU countries as regards the size and composition of immigrant population groups. Owing to labour market mechanisms, such groups are found mainly in the northern industrialized EU countries, whereas their presence in Mediterranean countries like Greece, Italy, Portugal, and Spain is rather limited. Mediterranean groups immigrate mainly to France or Germany. Portuguese, Spanish, and Maghreb residents concentrate in France, whereas Italian, Greek, former Yugoslavian, and Turkish residents concentrate in Germany. The largest immigrant groups in EU countries are Turkish and Maghreb residents; the latter originate from Morocco, Algeria, or Tunisia. Table 1.1 gives estimates of their size in twelve EU countries in January 1994. Table 1.1 Estimated numbers of inhabitants of Maghreb and Turkish origin in twelve EU countries, January 1994, based on the nationality criterion (EuroStat 1997) Maghreb countries Morocco Algeria Tunisia Total Maghreb Turkey Belgium 145.363 10.177 6.048 161.588 88.302 Denmark 3.180 368 404 3.952 34.658 Germany 82.803 23.082 28.060 133.945 1.918.395 Greece 333 180 314 827 3.066 Spain 61.303 3.259 378 64.940 301 France 572.652614.207206.336 1.393.165 197.712 Italy 77.180 3.177 35.318 115.675 3.656 The Netherlands 164.567 905 2.415 167.887 202.618 Portugal 221 53 28 302 65 Finland 560 208 142 910 995 Sweden 1.533 599 1.152 3.284 23.649 Great Britain 3.000 2.000 2.000 7.000 41.000 Total 1.112.695658.215282.595 2.053.505 2.514.417 According to EuroStat (1997) and based on the conservative nationality criterion, in 1993 the largest Turkish and Maghreb communities could be found in Germany (almost 2 million) and France (almost 1.4 million), respectively. Within the EU, the Netherlands is in second place as the country of immigration for Turkish and Moroccan residents. Given the decreasing significance of nationality and birth-country criteria, collecting reliable information about the composition of immigrant population groups in EU countries is one of the most challenging tasks facing demographers. Complementary or alternative criteria have been suggested in various countries with a longer immigration history, and, for this reason, a history of collecting census data on multicultural population groups. In English-dominant countries such as the USA, Canada, and Australia, census questions have been phrased in terms of self-categorization ('To which ethnic group do you consider yourself to belong?') and home language use. In Table 1.2, the four criteria mentioned are discussed in terms of their major advantages and disadvantages (see also Extra 1996; Broeder and Extra 1998a).

< previous page

page_3

next page >

< previous page

page_4

next page > Page 4

Table 1.2 Criteria for the definition and identification of population groups in a multicultural society (P/F/M = person/father/mother) Criterion Advantages Disadvantages Nationality • objective • (intergenerational) erosion (NAT) through naturalization or • relatively easy to (P/F/M) double NAT establish • NAT not always indicative of ethnicity/identity • some (e.g., ex-colonial) groups have NAT of immigration country Birth-country • objective • intergenerational erosion (BC) through • relatively easy to (P/F/M) establish • births in immigration country • BC not always indicative of ethnicity/identity • invariable/deterministic: does not take account of dynamics in society (in contrast to all other criteria) Selfcategorization • touches the heart of the • subjective by definition: matter also determined by (SC) • emancipatory: SC takes language/ ethnicity of account of person's own interviewer and by the spirit conception of ethnicity/ of times identity • multiple SC possible • historically charged, especially by World War II experiences Home • HL is most significant • complex criterion: who language criterion of ethnicity in speaks what language to (HL) communication processes whom and when? • HL data are • language not always core cornerstones of value of ethnicity/identity government policy in • useless in one-person areas such as public households information or education As Table 1.2 makes clear, there is no single royal road to a solution of the identification problem. Different criteria may complement and strengthen each other. Given the decreasing significance of nationality and birth-country criteria in the European context, the combined criterion of self-categorization and home language use is a potentially promising long-term alternative. 1.2.2 The Netherlands as case study For the reasons mentioned before, it is not easy to give a complete and reliable overview of the actual size of the immigrant population in any of the European Union countries. Typical for the Netherlands, in contrast to neighbouring countries like Germany and Belgium, some immigrant groups have had Dutch nationality since birth (cf. Extra and Vallen 1997). These include all Antilleans

< previous page

page_4

next page >

< previous page

page_5

next page > Page 5

(> 90.000) and most of the Surinamese (> 260.000) who came to the Netherlands in the last few decades from former Dutch colonies in the Caribbean, and the so-called repatriates from the former Dutch East Indies (the present Republic of Indonesia; > 280.000), who arrived in the Netherlands after Indonesia's independence. Bearing in mind the biased character of all available nationality statistics, some longitudinal trends in the size and growth of indigenous and non-indigenous population groups during the last decade can nevertheless be derived from recent data published by the Dutch Central Bureau of Statistics (CBS 1995). In the period 1985-1994, about 25% of the increase in the population of the Netherlands (890.000) was the result of an influx of immigrants, even though the immigrants as a group constitute less than 5% of the total population (> 15 million). Within the immigrant population, the strongest proportional growth stems from the Turkish, Moroccan, Surinamese, and (former) Yugoslavian communities, and from people belonging to the CBS category other non-European countries (mainly refugees). The latter group and the Turks and Moroccans also show the most substantial increase in terms of absolute figures. The Dutch Ministry of the Interior has attempted to reduce the increasing erosion of statistics pertaining to immigrant minorities in the Netherlands by proposing the following three ethnic determinants in all municipal population statistics (cf. Fernandes Mendes 1991): (1) birth-country of person, father, and mother; (2) nationality of person, father, and mother; (3) selfcategorization. Obviously, the combined birth-country criterion only suffices for first- and second-generation groups. Furthermore, the (combined) nationality criterion has limited value because many immigrant minority groups have or will obtain Dutch nationality. The third criterion, self-categorization, led to many objections being raised by both minority and majority groups in the Netherlands because of its subjective loading, the possibility of multiple self-categorization, and the potential misuse of the data collected. Ultimately, parliamentary support was given to the Ministry of the Interior for a gradual introduction of the combined birth-country criterion in all municipal population statistics, although it was recognized that this criterion would lead to a diminishing identification of immigrant minority groups over time (cf. Dales 1992:17). It was also recognized that other criteria could be relevant for specific purposes or domains. Explicit reference in this context was made to the relevance of the home language criterion in the domain of education (cf. Dales 1992:12). Table 1.3 gives an overview of population groups in the Netherlands on January 1, 1992, based on several birth-country criteria and the nationality criterion, and derived from Martens, Roijen, and Veenman (1994).

< previous page

page_5

next page >

< previous page

page_6

next page > Page 6

Table 1.3 Population of the Netherlands based on different identification criteria (BCPMF=combination of birth-country person, mother and father; NAT=nationality; BCP=birth-country person; BCM=birth-country mother; BCF=birth-country father) on January 1, 1992 (Martens et al 1994) Absolute figures Index (column 1 = 100) BCPMF NAT BCP BCM BCF Groups 12.764.767 113 Dutch 108 104 105 240.810 Turkish 89 66 96 99 195.536 Moroccan 84 67 95 99 262.839 Surinamese 8 65 87 86 90.650 Antillean 71 69 63 10.369 Greek 50 59 62 84 32.818 Italian 52 46 51 88 27.117 Former Yugoslavian 56 63 86 82 12.587 Portuguese 69 68 81 88 29.046 Spanish 58 61 75 81 14.330 Cape Verdian 19 65 99 96 5.631 Tunisian 46 56 66 94 39.762 Chinese 17 61 91 97 10.435 Vietnamese 46 83 99 96 1.392.435 18 Other groups 47 69 58 2.364.383 31 Total non-Dutch 54 77 71 15.129.150 100 Total Netherlands 100 100 100 Table 1.3 shows the significant differences which result when different identification criteria are applied. The combined birthcountry criterion in the second column shows a proportion of citizens of foreign origin of more than 15% (2.364.383 out of 15.129.150 persons). When the absolute figures in the second column are substituted by an index value of 100, it becomes clear that the 'Dutch' group is overrepresented on the basis of the nationality criterion and the BCP/BCM/BCF criteria, whereas all 'non-Dutch' groups are underrepresented on the basis of these criteria. For groups other than those mentioned in Table 1.3, only global estimates can be given and it is not clear what criteria are used for these estimates. Roelandt, Roijen, and Veenman (1991:31) give estimates from 1987/1988 for Moluccans (35.000), refugees (18.000), political refugees seeking asylum (8.350) and gypsies (3.700). From the more than 197.000 requests for asylum, about 57.000 were granted in the period 1985-1994 (CBS 1995). Table 1.4 gives an overview of population groups in the Netherlands on January 1, 1996, based on the combined birth-country criterion (BCPMF) versus the nationality criterion, and derived from recent CBS statistics (CBS 1997).

< previous page

page_6

next page >

< previous page

page_7

next page > Page 7

Table 1.4 Population of the Netherlands (x 1000) based on the combined birth-country criterion (BCPMF) and the nationality criterion on January 1, 1996 (CBS 1997) BCPMF Nationality Abs. diff. Groups (xlOOO) 12.872 14.768 1.896 Dutch 272 154 118 Turks 225 150 75 Moroccans 282 15 267 Surinamese 94 94 Antilleans 11 5 6 Greeks 32 17 15 Italians 56 34 22 Former Yugoslavs 13 9 4 Portuguese 29 17 12 Spaniards 17 2 15 Cape Verdians 6 2 4 Tunisians 1.585 331 1.254 Other groups 15.494 15.494 Total Compared to Table 1.3 and based on the combined birth-country criterion, Table 1.4 first of all shows an increase of the five largest immigrant minority groups over time (i.e., Turks, Moroccans, Surinamese, Antilleans and former Yugoslavs) and a stabilization of most other groups. Moreover, Table 1.4 again shows strong criterion effects of birth-country versus nationality. All immigrant minority groups are in fact strongly underrepresented in nationality-based statistics. As mentioned before, however, the combined birth-country criterion does not solve the identification problem. The use of this criterion leads to non-identification in at least the following cases: • an increasing group of third and further generations (cf. the Moluccan and Chinese communities in the Netherlands; see also Chapter 2); • different ethnocultural groups from the same country of origin (cf. Turks versus Kurds from Turkey); • the same ethnocultural group from different countries of origin (cf. Chinese from China versus Vietnam); • ethnocultural groups without territorial status (cf. gypsies). Verweij (1997) made a short tour d'horizon in four European Union countries (i.e., Belgium, Germany, France, Great Britain) and in the USA in order to study criteria utilized in the national population statistics of these countries. In Belgium, Germany, and France, such statistics have traditionally been based on the nationality criterion; only in Belgium has additional experience been gained with the combined birth-country criterion of persons, parents, and even grandparents. For various reasons, identification on the basis of the grandparents' birth-country is very problematic: four additional sources of evidence are needed (with multiple types of outcomes) and the chances of non-response are rather high. Verweij (1997) also discussed the experiences with the utilization of ethnic self-categori-

< previous page

page_7

next page >

< previous page

page_8

next page > Page 8

zation in Great Britain and the USA, leaving the home language criterion out of consideration. Given the increasing identification problems with the combined birth-country criterion, Verweij, on the basis of Anglo-Saxon experiences, suggested including the self-categorization criterion in future Dutch population statistics as the second-best middle- and long-term alternative in those cases where the combined birth-country criterion would not suffice. Moreover, he proposed carrying out small-scale experimental studies on the validity and social acceptance of the self-categorization criterion, given its subjective and historically charged character, respectively (see also Table 1.2), before this criterion would be introduced on a nation-wide scale. As early as 1982, the Australian Institute of Multicultural Affairs recognized the above-mentioned identification problems for inhabitants of Australia and proposed including questions on birth-country (of person and parents), ethnic origin (based on selfcategorization), and home language use in their censuses. In the next section, we will deal with such experiences abroad. 1.3 Census Experiences in English-Dominant Immigration Countries In this section, we focus on four non-European English-dominant immigration countries with long-term experiences in gathering nation-wide statistics on the multicultural composition of population groups in periodical censuses, i.e., Australia, Canada, The United States, and South Africa. In each of these countries, English has become the language of status and power as a result of colonization processes. At the same time, these countries are characterized by both indigenous and non-indigenous (immigrant) population groups that make use of other languages at home. We will deal in particular with the identification of ethnocultural minority groups by means of census questions about ethnicity and (home) language use. 1.3.1 Australia Every 10 years, Australia conducts a general census with the cooperation of the Australian Bureau of Statistics. On census day, forms, written in English and sent out in advance, are collected door-to-door by specially appointed census officials. Here we compare the general inquiries carried out in 1976 and 1986. Questions about language were incorporated in both inquiries: in 1976, the questions concerned family members of five years of age and older; in 1986, all family members. In Australia, no comprehensive data are available on language use before 1976, although as early as 1933 the census collected information about the ability to read and write in languages other than English (cf. Klarberg 1982). The 1976 census contained the following question about language (mentioning English, Italian, Greek, German, and Other): 'Indicate all languages regularly

< previous page

page_8

next page >

< previous page

page_9

next page > Page 9

used at home, at work, at school, when shopping, etc. Remember, many people may use more than one language - tick each language used regularly. If an Aboriginal language is used, (...) write in the name of the language.' The phrasing in terms of regular use was preferred to a possible phrasing in terms of language proficiency, mother tongue, or home language. An attempt was made to gather the broadest information possible about potential sources and needs with respect to languages other than English, and to avoid individual variation with respect to the self-evaluation of language proficiency and the self-interpretation of home language (for example, the individual's own versus his or her parents' home language). Moreover, a restriction to one particular domain (in this case, the home) would detract from the importance of language use in other circumstances. In decreasing order of frequency, the following languages were mentioned most often after English: Italian, Greek, German, Serbian/Croatian, French, Dutch, Polish, Arabic, Spanish, and Maltese. As regards numbers of speakers, only rough estimates could be given for some languages, owing to errors in computer processing and/or variation in the interpretation of the questions. Many parents indicated that their children 'regularly' used French, German, or Italian, although their contact with these languages was restricted to the classroom. The degree to which such language use was overrated can be gathered from the children's or the parents' places of birth. For predictive purposes, especially regarding expected processes of language shift and maintenance, data on home language use are more useful than data on regular use. The foundations for the acquisition of languages for intergenerational use are laid at home; languages learned at home stand the best chance of being maintained on the outside. Therefore, the 1986 census included the following language question for all the inhabitants of Australia: 'Does the person speak a language other than English at home? (No: speaks only English. If yes: please specify the language spoken).' In addition, the following question was asked: 'How well does the person speak English?'; the answer had to be given on a four-point scale: 'very well / well / not well / not at all'. The focus on home language use in the 1986 census makes the comparison with the 1976 data a difficult enterprise. Longitudinal research on census data about language use requires consistency in the phrasing of the questions, but such consistency is seldom available. The most important findings of the 1986 census were the following: • in total, 66 languages were mentioned, including English and all Aboriginal languages; • 14% of the Australian population said that they spoke a language other than English at home; • 12 languages were spoken at home by more than 50.000 people; • Italian and Greek were the languages most often mentioned, with 415.765 and 227.472 speakers respectively.

< previous page

page_9

next page >

< previous page

page_10

next page > Page 10

For a discussion of these and other findings, the reader is referred to Clyne (1991). When the 1976 census data are compared with those of 1986, a (sharp) increase in use is observed for some languages. This increase can be attributed to objective factors (like recent immigration) as well as subjective factors. In Australia, there has been a revival of the value and importance of ethnic origin (back-to-roots movement) and a concomitant revaluation by the government of the multicultural character of society. As a result, the use of languages other than English at home was rated more positively in 1986 than it had been in 1976, with corresponding consequences for the outcome of the surveys (cf. Fishman et al 1985, for comparable effects in the USA). For three languages, German, Italian, and French, the scores were lower in 1986 than in 1976 (cf. Clyne 1991:48-49 for an interpretation). The census data thus collected make further analyses of distributional features possible in terms of, among other things, regional spread, generation, age, sex, and socio-economic status (cf. Kipp, Clyne and Pauwels 1995). Such research, which involves the analysis of demographic data on language use repeatedly obtained on a large scale, is known as 'demolinguistics' and has developed into a flourishing discipline, especially in English-dominant immigration countries (cf. De Vries 1989). Data on the distribution over age groups, for example, make it possible to predict future changes in the ethnolinguistic composition of the population. In the 0-9 age category, both Italian and Greek have been overtaken in Australia by Arabic, while in the over-fifty age group, Italian and German are still in first and second positions. By cross-tabulating data on home language and place of birth, duration of stay or age, processes of language maintenance versus language shift (in the direction of English) can also be investigated, including the differences between and within various ethnic groups and generations. The data from the 1986 census show that the main factors in language maintenance versus shift relate to the spreading of groups, sex, and the degree of intra- versus interethnic marriages. Additionally, the extent to which the language in question forms a core value of ethnocultural identity for the group under consideration is very important (cf. Smolicz 1980, 1992). The smallest shift in the direction of English was observed in Mediterranean population groups for whom the home language has an important symbolic religion-related function, i.e., Moslems and Eastern-Orthodox Catholics. The strongest and fastest shift in the direction of English is taking place among people of Dutch origin (cf. Nieuwenhuysen 1993, for a wealth of interesting census data about the status of the Dutch community in Australia). Similar data about Dutchmen have been found in the USA and Canada. In the given English-dominant immigration contexts, Dutch does not seem to represent a core value of ethnocultural identity for many Dutchmen (see Klatter-Folmer and Kroon 1997, for counterevidence on the status of Dutch overseas). For a discussion of the factors that influence the intra/intergenerational speed of processes of language shift in Australia, the reader is referred to Clyne (1991, 1982), Romaine (1991), Ozolins (1993), and Kipp et al (1995).

< previous page

page_10

next page >

< previous page

page_11

next page > Page 11

Finally, remarkable differences in the phrasing of questions can be observed in nearby New Zealand, which also has a tradition of periodical population research. The 1991 census did not include a single language question, but did include the following questions about ethnic origin: • Which ethnic group do you belong to? New Zealand European, New Zealand Maori, Samoan, Cook Island Maori, Tongan, Niuean, Chinese, Indian, Other (please state); • Have you any New Zealand Maori ancestry? No/Don't know/Yes; • If yes, what is the main iwi (tribe) you belong to? Don't know/Don't belong to any iwi (tribe). • What other iwi (tribes) do you have strong ties with? (please state no more than two iwi). The instructions stipulated that in the first self-categorization question, more than one ethnic group could be mentioned. 1.3.2 Canada The first population study in Canada was conducted more than a century ago, in 1871. Until 1961, an official called on every home once every ten years, to interview the head of the family. Since 1971, a new procedure has been in use, which involves questionnaires (with extensive instructions), which it is the head of the family's responsibility to fill out and send back. Over the years, there have been various questions about ethnicity and language; the first important one was about ethnic origin. Before 1951, the phrasing was: 'To what ethnic or cultural group did you or your ancestor (on the male side) belong on coming to this continent?' In the instruction manual for the interviewer, it was explained, from 1901 onwards, that terms like 'Canadian' or 'American' do not refer to race and therefore must not be used. Examples were also given of terms that do refer to race. In addition, the manual contained the warning that nationality is not a foolproof indicator of ethnic origin, but that the language that was or is being spoken may be considered as such. In 1981, the principle that respondents could only mention one ethnic origin (namely, male descent) was abandoned; since then, they can also indicate whether they consider themselves to be of mixed origin, which, in view of the number of mixed marriages, is quite conceivable. De Vries (1985) points out that the question on ethnic origin does not yield thoroughly reliable data. Questions that require respondents to report on their own ethnic origin tend to measure either self-categorization or descent. Only if there is a strong correlation between the two, and if members of various ethnic groups do not marry outside their own group, does the item measure both components. Therefore, the above-mentioned question from the Canadian census is problematic in a number of respects.

< previous page

page_11

next page >

< previous page

page_12

next page > Page 12

In the 1901 census, two language items made their first appearance: a question about oral proficiency in Canada's two official languages, i.e., English and French, and a question about the respondent's mother tongue. Up to 1961, the first question was phrased as follows: 'Are you able to speak English, French or both well enough to carry on a conversation? (English only/ French only/ both/ neither.)' In the 1971 census manual, this question was clarified as follows: 'This conversation must be of some length and about several subjects.' Despite the fact that this clarification can be interpreted in various ways, there are indications of a strong correlation between reported and actual proficiency. The second question, regarding the respondent's mother tongue, was phrased as follows until 1941: 'What is the respondent's mother tongue?' The notion 'mother tongue' was defined as follows: '... the language first learned in childhood and still spoken.' From 1941 until 1971, the question concerned 'the language first spoken and still understood,' and in 1981, 'the language first learned and still understood.' In the last formulation, the possibility of a respondent 'forgetting' his/her mother tongue is taken into account. It can happen that a respondent learned a language as a child, but at the moment of the census no longer speaks (1901-1931) or understands (from 1941) it. In such cases, the respondent is asked to mention the next language that is still spoken or understood. Up to and including the 1971 census, respondents who had a bilingual upbringing had to make a choice between two mother tongues, since only one could be mentioned. The assumption was that their choice would be arbitrary. Research has shown, however, that the language with the greatest prestige was chosen (De Vries 1985). Beginning in 1976, respondents had the option to name two languages as their mother tongue, which increases the reliability of the data (De Vries and Vallee 1980). Since 1986, these multiple responses have been published as well. A final question about language in the Canadian census had to do with language use in the home environment. In 1971, this item was included on the recommendation of the Royal Commission on Bilingualism and Biculturalism (RCBB). However, the formulation 'What language do you most often speak at home now?' is ambiguous, since both you and the French vous fail to make clear whether the respondent is being addressed or his/her entire household. In the 1981 census, this ambiguity was rectified by the addition of yourself and vous-même respectively. In the 1967 RCBB report, in which the proposal to add the question about home language use was made, it was remarked that this item might have a great future. Provided that it would be useful in collecting valid data, 'home language' would become the basis for calculating the number of bilinguals in Canada. The criterion that had been applied up to that point, i.e., 'mother tongue,' would then be abandoned. Several researchers have qualified the data collected by means of the question on home language use as valid. However, no attempt has yet been made to change the criteria, and in the meantime, the RCBB has been dissolved.

< previous page

page_12

next page >

< previous page

page_13

next page > Page 13

In addition to the information gathered in the general population census, the General Social Survey (GSS), which has been held in Canada every year since 1985, also collects large-scale data. For this survey, the Canadian Statistics Bureau takes a random sample of 10.000 people of 15 years and older. A GSS consists of a core and a focus covering a different theme every year. The second GSS was focused on 'language,' and included a large number of questions on the subject. The interviewers presented a comprehensive questionnaire to the respondents. The structure of the section on language use was as follows. A respondent was asked to give his or her 'main language, that is, the language in which you are most at ease.' The answer could be one language or any possible combination of official and/or non-official languages in Canada. Subsequently, for each 'main language' mentioned, the skills and frequency of use were inquired about. Questions were also asked about knowledge and use of the two official Canadian languages, i.e., English and French, and about language use in the respondent's home environment during his or her childhood and adolescence. Language use at home, at school, and among friends was also addressed, as well as items on work and the working environment, oral and written language use at work, and communication with colleagues. Finally, there were questions about actual and preferred language use in government institutions, the media, and health care. It can be concluded that by means of these Canadian language inquiries, a comprehensive and detailed picture emerges periodically of language use in Canadian society. The inquiries yield material with which to complement and compare census data, thus laying a solid foundation for language research and language policy (see, for example, De Vries and Vallee 1980, Cummins and Danesi 1990). 1.3.3 The United States In the United States, there is a long tradition regarding the identification of ethnic minority groups. The importance and the utility of information about the language use of ethnic groups has been closely examined in numerous studies (Fishman 1989, Fishman et al 1985, Veltman 1983). Since 1790, population research has been conducted at 10-year intervals. Since 1902, the census has been carried out by a bureau specially founded for the purpose, the Census Bureau. Before 1960, officials from this organization called on every home to deliver a census questionnaire. This had to be filled out and handed to the official who came to collect it personally. Since 1980, the census forms have increasingly been sent by mail, with the request to fill them out and send them back (83% in the 1990 census), while in rural areas the forms have been delivered by an official. When questionnaires have not been completely filled in, officials of the Census Bureau pay a personal visit to the homes in question in order to gather the missing data. The census, which contains questions about race, language, origin, and ethnicity, consists of two kinds of questionnaires, a long one and a short one. The com-

< previous page

page_13

next page >

< previous page

page_14

next page > Page 14

prehensive questionnaire is intended for approximately 20% of the population. The short questionnaire contains questions about race and Spanish (Hispanic) origin. A random sample of the population is presented with the comprehensive questionnaire, which also contains questions about language. Even in the first census, in 1790, information was gathered about race on the basis of self-categorization. After each census, this question was evaluated and, if necessary, revised, because the formulation of the question turned out to strongly influence the answers. In the 1990 census, the wording of the question about race from the 1980 census was changed because research had shown that this question yielded data that were insufficiently precise. In 1980, the question had been 'Is this person ...,' followed by an enumeration of a number of races, plus a blank in which a different race could be filled in. It was not explicitly indicated, however, that the question concerned races. In the 1990 census, the idea was for the designation 'race' to be used, but with a restricted list of races. The 1980 census question contained a comprehensive enumeration of Asian races and races of islands in the Pacific. The intention was to combine these categories and give a number of examples. This plan met with fierce resistance from the groups in question, since it was feared that this question would result in a 'loss of identity.' Eventually, a compromise was worked out, in that with the category Asian or Pacific Islander, a comprehensive enumeration was printed of races that could be ticked, but also a race other than Asian or Pacific Islander could be filled out. Since 1970, the census also contains a question about Spanish/Hispanic origin: 'Is this person of Spanish/Hispanic origin?' (1990). The question was not combined with race, since it was feared that such a combination would result in an underestimation of a number of racial groups. In the 1980 and 1990 censuses, there is also, for the first time, a general question about ethnicity. Like the question about race, it is based on self-categorization. The ethnicity question, however, has been given an open formulation: 'What is this person's ancestry or ethnic origin?' The Census Bureau looks upon ethnicity and race as two different concepts: people belonging to a certain ethnic group can be of any race. Since 1910, the censuses have also contained questions about language, which yield a wealth of data and frequently form the basis for sociolinguistic research (e.g., Lieberson 1981). In the United States, two kinds of questions about language are asked in the census, i.e., those about the respondent's first language and those about the respondent's proficiency in English. The definitions of these notions show an enormous diversity in consecutive censuses and, therefore, exert a negative influence on the longitudinal comparability of the data collected. For example, in 1940, the question about mother tongue was: 'What language was spoken in this person's home when he was a child?' The 1970 question was quite similar to the 1940 one, but yielded totally different results: 'What language, other than English, was spoken in this person's home when he was a child?' (Veltman 1983). The addition 'other than English' should make it clear that it was a minority language that was being inquired about. As a result, the two language ques-

< previous page

page_14

next page >

< previous page

page_15

next page > Page 15

tions are not comparable, because the 1970 question may have been interpreted as 'Did you ever hear a language other than English spoken in your home when you were a child?' To find out to what extent the 1970 census question was given an interpretation different from that given to the 1940 one, a reinterview study conducted in 1971 (to check on data collected through the census) again asked respondents who had given an affirmative answer to the 1970 census question whether in their childhood they themselves: • were monolingual in the minority language in question; • mostly used the minority language, but also spoke English; • usually spoke English, but also regularly spoke the minority language; • sometimes spoke the minority language; • seldom spoke the minority language; • did not speak the minority language at all. It appeared that only 68% of the respondents who had mentioned a language other than English in answer to the 1970 census question actually had that language as their mother tongue. The 1970 data that were officially published by the Census Bureau show that the mother tongue of 16% of the population of the United States is a minority language. On the basis of the results of the reinterview study, however, this group was estimated at 8%. It encompasses all the people that in their childhood were monolingual in the minority language, mostly used the minority language besides English, or used the two languages equally often. From 1910 to 1970, the census contained questions about mother tongue. In 1980 and 1990, the language questions consisted of three parts: • does this person speak a language other than English at home? • what is this language? • how well does this person speak English? The emphasis in these questions was on current language usage, i.e., the language the respondent was using at that particular moment. The idea was to use the data collected by means of this question to assess the need for bilingual education and for other forms of bilingual services. The 1990 census data published by the Census Bureau in several separate reports show that almost a quarter of the nearly 20 million Americans who were born abroad, emigrated between 1985 and 1990. The main countries of origin were Mexico (4.3 million) and the Philippines (913.000). Canada, Cuba, Germany, Great Britain, Italy, South Korea, Vietnam, and China each contributed at least 500.000 inhabitants. The states with the highest number of inhabitants of foreign extraction were California (33%), New York (14%), Florida (8%), Texas (8%), New Jersey (5%), and Illinois (5%). In 1990, almost 32 million people (14% of the American population of 5 years of age and older) indicated that they spoke a language other than English at home, as compared with 23 million (11%) ten years earlier. After English, Spanish was

< previous page

page_15

next page >

< previous page

page_16

next page > Page 16

the language most frequently spoken at home (17.3 million), followed by French (1.7 million), German (1.5 million), Italian (1.3 million), and Chinese (1.2 million). Approximately 4.5 million people spoke an Asian or Pacific Islander language at home, and almost 332.000 people, a native North American language. Of the non-English-speaking population, 60% declared that they spoke English very well, and 20% that they did not speak English or spoke it badly. A relatively low command of English was reported mainly among Asian language groups. 1.3.4 South Africa Compared to the three countries discussed above, South Africa provides a very different picture, due to its particular spectrum of indigenous and non-indigenous languages and to its politically burdened history of apartheid. During the period of apartheid (1948-1994), English and Afrikaans were the only two languages with an officially recognized nation-wide status, despite the wide variety of other languages learnt and spoken in South Africa. Apart from Afrikaans, English, and other languages of European origin, two major groups of languages should be mentioned here, i.e., • Bantu languages, in particular (isi)Zulu, (isi)Xhosa, (si)Swati, (isi)Ndbele, (se)Sotho, (se)Tswana, (xi)Tsonga (tshi)Venda, and Sepedi; • Indian languages, in particular Hindi, Gujarati, Tamil, Urdu, and Telegu. While Bantu languages have their roots in Southern Africa, European and Indian languages originate from abroad, coming into South Africa since the 17th and 19th centuries respectively. For a historical and sociolinguistic discussion of the spectrum of languages in South Africa, we refer to Mesthrie (1995), Extra and Maartens (1998) and Broeder, Extra and Maartens (1998). In May 1996, the Constitutional Assembly of the post-apartheid Republic of South Africa adopted a new Constitution, which provides in clause 6 for no less than eleven official languages: (1) The official languages of the Republic are Sepedi, Sesotho, Setswana, siSwati, Tshivenda, Xitsonga, Afrikaans, English, isiNdebele, isiXhosa and isiZulu. (2) Recognising the historically diminished use and status of the indigenous languages of our people, the state must take practical and positive measures to elevate the status and advance the use of these languages. (3a) The national government and provincial governments may use any particular official languages for the purposes of government, taking into account usage, practicality, expense, regional circumstances and the balance of the needs and preferences of the population as a whole or in the province concerned; but the national government and each provincial government must use at least two official languages.

< previous page

page_16

next page >

< previous page

page_17

next page > Page 17

(3b) Municipalities must take into account the language usage and preferences of their residents. (4) The national government and provincial governments, by legislative and other measures, must regulate and monitor their use of official languages. Without detracting from the provisions of subsection (2), all official languages must enjoy parity of esteem and must be treated equitably. (5a) A Pan South Africa Language Board established by national legislation must promote and rate conditions for the development and use of: (i) all official languages; (ii) the Khoi, Nama, and San languages; and (iii) sign language; and (5b) promote and ensure respect for: (i) all languages commonly used by communities in South Africa, including German, Greek, Gujarati, Hindi, Portuguese, Tamil, Telegu, and Urdu; and (ii) Arabic, Hebrew, Sanskrit, and other languages used for religious purposes in South Africa. In the changing South Africa of the nineties, language planning has become a primary area of debate on the national agenda of reform. In this context, reliable census data on (home) language use are referred to as prerequisite. Censuses have been held in South Africa during this century at intervals of ten years and, since 1991, at five-year intervals. The 1980 and 1991 census data are the last and first ones in the apartheid and post-apartheid era respectively. Information based on the 1980 census data is provided by Grobler, Prinsloo, and Van der Merwe (1990). The 1991 census data have been documented by Luüs and Oberholzer (1994), and Krige et al (1994). Van der Merwe and Van Niekerk (1994) provide most interesting comparative data on the 1980 and 1991 censuses in their Language Atlas of South Africa. Both the 1980 and 1991 censuses were based on questionnaires, written in English and Afrikaans only, and delivered to the heads of households. Both censuses contained questions on ethnicity/race and language. In the 1991 census, the first question asked for 'population group' in terms of 'White/Coloured/Asian/Black.' The second question was formulated as follows: 'Indicate whether each person (in the household) can speak (communicate in), read and/or write the following languages: Afrikaans/English/Black Language/Other.' In addition, the following two questions were asked: 'State which language each person most often speaks at home' and 'If more than one language is usually spoken at home, state the other language which is spoken.' Black languages should be specified in terms of the Bantu languages distinguished before. In the most recently held 1996 census, available in any of the eleven official languages, the phrasing on ethnicity/race was: 'How would (the person) describe him-/herself?' Possible answers related to 'African/Black, Coloured, Indian/Asian, and White.' No further specifications were given, nor was there room for

< previous page

page_17

next page >

< previous page

page_18

next page > Page 18

other specifications. The two questions regarding language use were phrased as follows (census 1996): 'Which language does (the person) speak most often at home?,' and 'Does (the person) speak more than one language at home?' The answer to the former question had to be specified in an open space (no elucidation was given), while the answer to the latter question had to be specified as yes or no. If yes, 'The language (the person) speaks next most often' was asked for. The relevant census data in all cases were supplied in report form by the Central Statistical Service (CSS) as provided by the South African government. As yet, no data have been made available on the outcomes of the 1996 census. Van der Merwe and Van Niekerk (1994) provide comparative data on dominant home languages, as referred to in the censuses of 1980 and 1991. Their Language Atlas of South Africa includes language maps of Zulu, Afrikaans, Sesotho, saLeboa, English, Xhosa, Sesotha, Tsonga, Tswana, Swati, Ndbele, Venda, European immigrant languages, and Oriental languages. In all cases, the distribution of dominant home languages was specified by province, according to the 1980 and 1991 census data respectively, and maps were provided for the pattern of language change between 1980 and 1991. According to the 1991 census data, Zulu (8.3 million people, 22% of South Africa's total population), Xhosa (6.6 million, 18%), and Afrikaans (5.7 million, 15%), dominate the linguistic scene. The other prominent languages are Sesotho, saLeboa, English, and Sesotho. The users of the nine African languages jointly constitute approximately two-thirds of the total population of the country. Oriental languages (in particular Tamil, Hindi, Telegu, Gujarati, Urdu, and Chinese) are used as a home language by only a small fraction of the population, as is the case with the European immigrant languages (in particular Dutch, French, German, Greek, Italian, and Portuguese). Analyses of dominant home language use in South Africa in 1980 reveal a clear pattern of regional dominance for each of the eleven languages that have enjoyed official status since 1994. Although Afrikaans was the dominant home language in the largest area of South Africa (roughly speaking, the South Western part), with the exception of the Cape Town and Pretoria metropolitan areas, this area is relatively thinly populated. On the other hand, English was the dominant language in only a few districts, in particular the Johannesburg and Durban metropolitan areas, but all these were densely populated. The 1991 census data reveal little change in the patterns of language distribution and language dominance. Some districts became Zulu versus Afrikaans dominant, probably due to processes of migration, while the Johannesburg area switched from English-dominant to Zulu-dominant. In spite of its face value, and owing to its history of apartheid, the census figures available for South Africa show a unique set of inadequacies (cf. Maartens, 1998). In South African censuses up to and including the 1980 census, the process whereby adjustment for undercount was made was completely lacking in transparency and no methodology was made available to the public. The published figures already incorporated the undercount adjustment. In the 1985 census reports, the unadjusted figures were provided together with the lists giving the

< previous page

page_18

next page >

< previous page

page_19

next page > Page 19

recommended adjustment by race, gender, and age. No provision was made for the incorporation of district-related (urban/rural) differences. After the 1991 census, validation and adjustment was coordinated by the BMR (Bureau of Market Research), UNISA (University of South Africa), and representatives from organizations such as the HSRC (Human Sciences Research Council) and the DBSA (Development Bank of Southern Africa), with the result that these figures are accepted as the most accurate and dependable of any South African census. Maartens (1998) reports, however, that the structural constraints inherent in the earlier censuses are apparent here as well. Language data from the former TBVC states (Transkei, Bophuthatswana, Venda, and the Ciskei) is unavailable and is simply either represented by an estimate in the distribution figures available or ignored, as in the 1980 figures. Moreover, it should be kept in mind that years of rapid political and social change have passed since the last census. Widespread urbanization and the influx of illegal immigrants, primarily from across the borders of South Africa, are factors to contend with on the language scene. In 1994, the territorial division of South Africa into four provinces, six homelands and the four independent TBVC states changed into a nine-province division - a fact to be kept in mind when considering pre-1994 language maps. It remains to be seen how the reported inadequacies will be treated in future censuses. A serious redress of these problems in combination with anticipated changes in the population distribution, however, might lead to a drastic modification of the pattern of language distribution and language dominance in South Africa. 1.4 Conclusions and Discussion. From the census experiences in the reviewed non-European English-dominant immigration countries, it becomes clear that nation-wide population data on ethnicity and (home) language use are valuable instruments for indicating the multicultural composition of societies. Language-related data can be, and in some countries are, used as the basis for national, regional, and/or local policy decisions in such domains as information or media services and educational planning. The review of English-dominant immigration countries shows that census questions on language and ethnicity have not only been phrased differently in the countries under consideration, but also in successive censuses. Owing to such differences, comparative cross-national and longitudinal studies of the available statistics are not easy undertakings. There is another methodological issue regarding the quality of census data. An important similarity between the census questions discussed in the previous sections is that all of them lead to reported assessments rather than to observed facts. It is a well-known phenomenon that various response effects can play a role in survey studies among minority groups. Meloen and Veenman (1990) make a distinction between the effects of the structure of the

< previous page

page_19

next page >

< previous page

page_20

next page > Page 20

questionnaire, the characteristics of both interviewer and respondent, and the interaction between these factors. The answer to questions about both ethnicity and language use can be coloured by the language in which the questions are asked (which may or may not be the language of the respondent), by the ethnicity of the interviewer (idem), by the intended and/or perceived aim of the research (which may or may not be an anonymous collection of data, and may or may not be commissioned by the government), and by the spirit of the times, which may or may not have a multicultural orientation. Moreover, problems of interpretation can arise with the identification and classification of languages on the basis of the answers given. For a discussion of these and other possible effects, the reader is referred to Nicholas (1992) and Alladina (1993). The problems mentioned are inherent in the method and scale of data collecting and can only partially be cushioned by small-scale empirical research into actual language use. Such research is no alternative to large-scale surveys by means of questionnaires, but constitutes a valuable complement to it. In reviewing the available cross-national data, we have come across remarkable differences in the denotation of the groups under discussion. In English-dominant immigration countries, minority groups are frequently referred to as 'non-English speaking groups' or curtly 'NES groups.' Obviously, such reference is an expression of English as the language of status and power, and it does not do justice to the languages these groups do or can speak. A notable difference between non-European English-dominant immigration countries and European Union immigration countries is that it is only in the former countries that immigrant minority groups have succeeded in replacing the languages of previous prestige (see McDermott 1998, for a discussion of the rise of English in South Africa). Also within the European Union, however, English is challenging the role of other prestigious national languages as the medium of intercultural communication. Another striking difference between the referential framework of statistical population data on EU countries and Englishdominant immigration countries emerges in reviewing the literature. While in the latter countries the concept of '(ethnic) minorities' is common language, EU statistics invariably refer to 'non-national residents,' 'foreign residents,' or simply 'nonnationals' or 'foreigners.' Parallel to such reference, immigrant minority languages in EU countries are often referred to as 'nonindigenous,' 'non-regional' or even 'non-territorial' languages. The conceptual exclusion rather than inclusion of immigrant minority groups and immigrant minority languages is clearly linked with a restrictive conceptualization of the notions of citizenship and nation-state. In non-European English-dominant immigration countries, these notions are commonly shaped by a constitutional ius solis (law of the ground) in terms of which nationality derives from the country of birth. Such notions in EU countries are commonly shaped by a constitutional ius sanguinis (law of the blood) in terms of which nationality derives from parental origins. Interestingly, ius sanguinis is maintained by many European countries when immigrants seek citizenship in them. Emigrants from European countries who settled abroad, however, have instituted ius solis in the

< previous page

page_20

next page >

< previous page

page_21

next page > Page 21

countries which they colonized, in order to legitimate their claim to citizenship. In establishing the constitutions of (sub)continents like the USA, Canada, Australia, and South Africa, no consultation whatsoever took place of such native groups as Indians, Eskimos, Aboriginals and Zulus, respectively. In this context it is a remarkable phenomenon that Canada and South Africa have in the past referred to themselves as 'bilingual' English-French and English-Afrikaans countries, respectively. As yet, little experience has been gained in European Union countries with periodical censuses, or, if such censuses have been held, with questions on ethnicity or (home) language use. A case in point is Great Britain (see also Chapter 3, Section 3.2.4). Ever since 1841, general population research has taken place in Great Britain once every 10 years (with the exception of 1941, and with the census in 1991). In 1991, the questions were posed in English, and a clarification was available in eleven different languages, including Welsh and Scottish. The identification of the various population groups took place on the basis of country of birth. The first time questions were asked about ethnicity was in 1991 (cf. White 1990, Haskey 1991). This concept was operationalized as: 'White/ Black-Caribbean/ Black-African/ Black, Other/ Indian/ Pakistani/ Bangladeshi/ Chinese/ Any other ethnic group.' Although plans had been made to include language questions with a view to the identification of population groups, in 1991, only the census in Wales and Scotland inquired about the level of proficiency in indigenous minority languages: 'Does the person speak/ read/ write Welsh?' and 'Does the person speak/ read/ write Scottish Gaelic?' A comprehensive discussion of the considerations that led to the inclusion or exclusion of questions about ethnicity in the English census can be found in Sillitoe (1987) and Sillitoe and White (1992). It is expected that, as a consequence of ongoing processes of immigration and minorization, European Union countries will show a development towards periodical censuses with questions on language and ethnicity. Given the decreasing significance of nationality and birth-country criteria, the combined criterion of ethnocultural self-categorization and home language use would be a potentially promising long-term alternative for obtaining basic information on the multicultural composition of societies.

< previous page

page_21

next page >

< previous page

page_ii

next page >

< previous page

page_23

next page > Page 23

2 Sociolinguistic perspectives: Case studies on the vitality of immigrant minority languages This chapter deals with the aims and outcomes of a study on the vitality of immigrant minority languages amongst primary school children in the Netherlands. The aims and method of the study are outlined in Section 2.1. Section 2.2 deals with the first phase of the project, which aimed at establishing the feasibility and validity of language-related criteria as determinants of ethnicity. Sections 2.3 and 2.4 focus on the second phase of the project. Here, the outcomes of a home language survey are presented in terms of the inventory and distribution of reported home languages and in terms of detailed profiles of eleven major language groups. Conclusions and a discussion of the study are presented in Section 2.5. 2.1 Aims and Method In the previous chapter (Section 1.2.2), demographic data were presented that related to the Netherlands as an emerging multicultural society. It was made clear that there is no census tradition in the Netherlands, let alone a tradition of collecting nation-wide statistics on ethnicity and/or (home) language use. As a result, little information is available on the multicultural composition of population groups. Although home language use has been recognized by the Dutch Ministry of the Interior (Dales 1990) as a complementary or alternative criterion of ethnic identity and as prerequisite for educational policy on first and second language instruction, this criterion has never been utilized in practice (compare in contrast, Chapter 3, Section 3.2.6. for nation-wide statistics on home language use of school pupils in Sweden). In the early 1990s, a project was carried out in the Netherlands which was directed towards gaining insight into the status of immigrant minority languages at home. The aims of the project were threefold: (1) to establish the feasibility and validity of collecting data on home language use of children as possible determinants of ethnicity; (2) to collect data about home language use on a larger scale than had been attempted before in the Netherlands; given the educational relevance of such data for home language instruction (see also Chapter 3, Section 3.2.5.), the target groups in the survey were primary school children; (3) to collect multidimensional data on language use, in contrast to the nationwide censuses reviewed in the previous chapter which contained only few language-related questions.

< previous page

page_23

next page >

< previous page

page_24

next page > Page 24

The project was carried out in two phases. First, a feasibility and validity study was carried out in 1992 among almost 6.000 primary school children in the city of 's-Hertogenbosch (see Broeder et al 1993, Broeder and Extra 1994, 1995b, Broeder, Extra and Van Hout 1995). In the second phase, an extensive home language survey was carried out of nearly 35.000 children in the south of the Netherlands (see Broeder and Extra 1995c, 1996a, 1997a, 1998b). 2.2 Language-Related Criteria as Determinants of Ethnicity The reported outcomes on language use of immigrant minority groups in the countries discussed in Chapter 1 derive from allpopulation censuses and/or large-scale surveys carried out and repeated at regular intervals. They provide interesting evidence on the size and status of minority groups in a multicultural society. Experiences in other countries reveal that the reliability and validity of the minority language profiles are restricted by at least the following factors (see also Chapter 1, Section 1.4, Nicholas 1992, and Alladina 1993): • The exact phrasing of the language-related questions shows large variation, both within and between countries. • Reported evidence on language use is most commonly based on self-judgements rather than observed facts. • Answers to questions on language use are potentially biased by a number of factors such as: the language in which the questions are asked; the ethnicity of the interviewer; the perceived objectives of the questions; the pre-coded listing of languages in the questionnaires; and the spirit of the times, which may or may not be favourable towards multiculturalism in society. The seriousness of these restrictions varies according to the method and the scale of data collection. Small-scale empirical studies on observed language behaviour may counterbalance bias effects by testing the validity of reported questionnaire-based data. Such small-scale studies are not an alternative to large-scale questionnaires, but they are a valuable complement. The combined collection of both reported and observed evidence on immigrant minority language use was aimed at in the first phase of the project that was carried out in the Netherlands in 1992. 2.2.1 Feasibility study An important aim in the first phase of the project was to investigate the feasibility of collecting data on home language use of school children in an efficient way, without putting too much of a burden on both the children and their teachers. The data were collected on a relatively large scale by using a relatively short questionnaire which had to be filled out by primary school teachers for all

< previous page

page_24

next page >

< previous page

page_25

next page > Page 25

children in their classes. Part of the questionnaire was inspired by the School Language Survey used in the Linguistic Minorities Project that was carried out in Great Britain (LMP 1985; see also Chapter 3, Section 3.2.4). After some background information was gained, the teacher asked the following screening question: •Is any language other than Dutch ever used in your home? Only those children who gave an affirmative answer, were asked language-related questions for the following domains: • Language Which languages are used in your home? variety: • Language Can you understand/speak/read/write this language? proficiency: • Language Do you speak this language with your mother/father/older choice: brother(s) or sister(s)/younger brother(s) or sister(s)? • Language Which language do you speak best? dominance: • Language Which language do you like to speak most? preference: In addition, the teachers collected the following information from all the children in their classes: • Self-categorization: To which group do you consider yourself to belong? • Birth-country:

of the child, mother, and father.

• Nationality:

of the child, mother, and father.

The feasibility study was carried out in 's-Hertogenbosch, a city in the south of the Netherlands with a population of about 90.000. The relative size and variety of immigrant minority groups in 's-Hertogenbosch mirrors the composition of the Dutch population as a whole quite well. Data were collected on 5.788 children from 31 primary schools; this is about 75% of the total primary school population of 's-Hertogenbosch. The data reflect a representative sample with respect to suburb, age, school type (e.g., in terms of religious denomination), immigrant group, and gender of the children. A subsample of 1.272 children (about 22%) could be identified on the basis of at least one of the following criteria for immigrant minority group membership: birthcountry or nationality of the child, father, or mother, and home language use other than Dutch. Within this subsample, five major immigrant minority groups could be distinguished, i.e., 428 Moroccan children, 190 Turkish children, 146 Surinamese children, 56 Antillean children, and 47 Moluccan children. Moroccan and Turkish children will be referred to as the Mediterranean groups; Surinamese, Antillean, and Moluccan children as the ex-colonial groups. The latter groups were distinguished because Dutch has (had) a special status in the former Dutch colonies; the high status of Dutch and/or the special position of Dutch in the primary educational system in these countries may have (had) a positive effect on the Dutch language skills of these children. A substantial group of 405 children had a wide diversity of ethnocultural backgrounds.

< previous page

page_25

next page >

< previous page

page_26

next page > Page 26

The adequacy of the traditional ethnic identification criteria can be established on the basis of the answers recorded by the teachers for their children. Table 2.1 shows the number of children identified through each of the non-language criteria. Table 2.1 Ethnic identification on the basis of six non-language criteria (*combined criterion of birth-country of child/mother/father) Ethnic Nationality Birth-country of Combined Selfgroup child child mother father birthcategorization country* child Moroccan 339/404 121/404 381/416 387/402 404/414 242/308 N=428 84% 30% 92% 96% 98% 79% Turkish 150/176 38/176 168/180 174/177 178/181 117/149 N=190 85% 22% 93% 98% 98% 79% Surinamese 13/134 14/142 111/138 108/129 132/139 31/108 N=146 10% 10% 80% 84% 95% 29% Antillean (0/36) 34/55 46/55 38/47 54/55 20/44 N=56 0% 62% 84% 81% 98% 45% Moluccan 2/40 2/44 20/43 19/39 27/38 10/37 N=47 5% 5% 47% 49% 71% 27% The first column in Table 2.1 reflects the total number of children within each immigrant minority group. The table shows that for certain children, information on a particular criterion is missing. Identification through nationality shows a high degree of missing data for the Antilleans. Most Antillean children are of Dutch nationality. Nevertheless, this information was not recorded for 20 of the 56 children. With respect to the identification through self-categorization, the amount of missing data is also remarkably high. Teachers may have had negative feelings about this criterion and/or children, particularly by younger children, may have had problems with self-categorization. Table 2.2 reflects the coverage of the different identification criteria within each group. Table 2.2 Coverage of non-language criteria: 0-10%=--, 10-30%=-, 30-70%=0, 70-90%=+, 90-100%=++ Ethnic Nationality Birth-country of Combined Selfgroup child child mother father birthcategorization country* child Moroccan + ++ ++ ++ + Turkish + ++ ++ ++ + Surinamese --+ + ++ Antillean -0 + + ++ 0 Moluccan --0 0 + -

< previous page

page_26

next page >

< previous page

next page >

page_27

Page 27 A clear pattern emerges from Table 2.2. The nationality criterion is insufficient: only for the Moroccan and Turkish group could a fair number of children be identified. The child's country of birth shows a low coverage for all groups because a large number of children belong to the second generation. The combined birthcountry criterion (i.e., either the child or one of the parents was not born in the Netherlands) seems to present a fairly complete picture. However, the low coverage for the Moluccan group indicates that this combined criterion will also be problematic for the identification of third and next generations. Finally, the criterion of self-categorization does not seem to be an adequate alternative for nationality and country of birth. The coverage of identification through self-categorization is limited and only a small number of children can be traced within the Surinamese, Antillean, and Moluccan groups. For the language profile of each child, five language-related criteria were applied which might have a potential value in the identification of ethnic group membership: • Home language use: any other language than Dutch used at home. • Language proficiency:

the child is reported to understand this language (other than Dutch).

• Language choice:

the child is reported to often speak this language (other than Dutch) with its mother.

• Language dominance:

the child is reported to be more or equally fluent in this language (other than Dutch).

• Language preference:

the child is reported to have an equal or stronger preference for this language (other than Dutch).

Table 2.3 shows the number of children identified within each group through each of these five language-related criteria. Table 2.3 Ethnic identification on the basis of five language-related criteria Ethnic Home Language Language Language group language proficiency choice dominance Moroccan 416/428 398/422 316/422 223/418 N=428 97% 94% 75% 53% Turkish 185/190 174/186 144/185 101/183 N=190 97% 94% 78% 55% Surinamese 108/144 84/135 10/128 9/137 N=146 75% 62% 8% 7% Antillean 42/55 37/53 21/51 19/54 N=56 76% 70% 41% 35% Moluccan 42/47 31/46 10/44 4/44 N=47 89% 67% 23% 9%

< previous page

page_27

Language preference 218/412 53% 100/181 55% 28/133 21% 17/52 33% 14/41 34%

next page >

< previous page

page_28

next page > Page 28

The outcomes in Table 2.3 show that the amount of missing data is much lower for the language-related criteria than for the earlier mentioned non-language criteria. The coverage of the language identification criteria within each group is shown in Table 2.4. Table 2.4 Coverage of language-related criteria: 0-10%=--, 10-30%=-, 30-70%=0, 7090%=+, 90-100%=++ Ethnic Home Language Language Language Language group language proficiency choice dominance preference Moroccan ++ ++ + 0 0 Turkish ++ ++ + 0 0 Surinamese + 0 --Antillean + + 0 0 0 Moluccan + 0 -0 Table 2.4 shows that the criterion of home language use has the highest coverage, in particular for Moroccan and Turkish children. With respect to the criteria of language proficiency and language choice, an interesting difference emerges between the Mediterranean and ex-colonial groups. Proportionally, more children within the Turkish and Moroccan groups report that they understand the language other than Dutch spoken at home and that they speak this language with their mother. Important additional information is provided by the criteria of language dominance and language preference. As can be derived from Table 2.3, the percentages for reported language dominance and language preference (i.e., other than Dutch) have a comparable level within the Moroccan, Turkish, and Antillean groups. However, within the Surinamese and Moluccan groups, relatively few children report dominance of the other language, compared to the number of children who report that they prefer the other language. A comparison of the coverage by the language criteria and the non-language criteria shows that the best determinants of ethnicity are home language use and the combined birth-country criterion, respectively. Of course, the success of the latter criterion will by definition decrease as the number of third and next generation children increases. The complementary value of the two criteria can be established by extracting those children who are only covered by one of the criteria from the sample. The number of children with different outcomes on the criteria of home language use and combined birth-country is presented in Table 2.5.

< previous page

page_28

next page >

< previous page

page_29

next page > Page 29

Table 2.5 Number of children with different outcomes on the criteria of home language use and combined birth-country Identification criteria Moroccan Turkish Surinamese Antillean Moluccan Comb. birth-country non-Dutch 11 4 36 13 11 Home language Dutch Combined birth-country Dutch 10 3 7 1 10 Home language non-Dutch Total 21 7 43 14 21 Table 2.5 clearly shows that the combined birth-country criterion cannot simply be replaced by the criterion of home language use. In addition, it is obvious that the criterion of home language use traces children who are outside the scope of identification through the combined birth-country. This holds in particular for Moluccan children. Of all the groups under discussion, Moluccans have the longest migration history in the Netherlands (see also Section 2.4.8). 2.2.2 Validation study An additional aim of the first phase of the project was to provide insight into the validity of the reported ethnic criteria, in particular the validity of the reported language data. The focus of the validation study was on the two largest groups, i.e., the Turkish and Moroccan children. In both cases, children were selected who were at the start (grades 1/2, commonly ages 4-6) or at the end (grades 7/8, commonly ages 10-12) of their primary school career. The Turkish group consisted of 67 children (32 young and 35 older children). The Moroccan group consisted of 141 children. The Moroccan children were oversampled in order to include enough children to allow for a differentiation between Arabic-speaking children (49 young and 33 older children) and Berber-speaking children (26 young and 33 older children). In other studies, Surinamese children (Habraken 1992) and Moluccan children (Strating 1994) were taken into account. The children were interviewed individually in Turkish, Arabic, or Berber by a native speaker of each of these languages. The interviewer used an extensive questionnaire with 89 questions. In addition to the questions from the survey study, this questionnaire also included more detailed questions on communication patterns and on language attitudes. If the child had problems understanding the questions in the immigrant minority language, the interview was repeated in Dutch a few days later by a native speaker of Dutch. In addition to the questionnaire, a bilingual receptive vocabulary test was administered in order to collect observed instead of reported data on the proficiency of the children in the pertinent minority language and in Dutch. The vocabulary test developed by Verhoeven et al (1995) was used for the youngest children in grades 1/2. There were three versions of the test: a Turkish version for the Turkish children, an Arabic version for the Moroccan children, and a Dutch version

< previous page

page_29

next page >

< previous page

page_30

next page > Page 30

for both groups. The test consisted of 60 items (multiple choices of content words). The receptive vocabulary test developed by Aarssen, De Ruiter and Verhoeven (1992) was used for the older children in grades 7/8. Again, there was a Dutch, Turkish, and Arabic test version. The test consisted of 36 items (content words). A native Turkish/Moroccan interviewer administered the Turkish/Arabic version of the test. A few days later, a Dutch interviewer administered the Dutch version of the test. The most direct way to discover a possible bias or instability in the data is to compare the responses of children in the feasibility study with the responses of the same children to the same questions in the validation study. In this comparison, two factors were considered: the number of children for which the pertinent information is missing in one of the studies and the number of children for which different responses have been recorded. The results of this comparison of identification criteria are summarized in Table 2.6. Table 2.6 Number of missing values and number of shifts in comparing the ethnic identification criteria in the feasibility study and the validation study Identification criteria Turkish children Moroccan children missing shifts missing shifts Nationality 8 8 14% 16 18 14% Birth-country child 7 6 10% 9 29 22% Birth-country mother 6 0 0% 10 7 5% Birth-country father 7 0 0% 13 6 5% Combined birth-country 4 1 1% 8 6 5% Self-categorization 13 9 17% 43 16 16% Home language 0 0 0% 0 1 1% Language proficiency 2 3 5% 1 3 2% Language choice 2 9 14% 4 29 21% Language dominance 5 7 11% 3 45 33% Language preference 6 17 28% 5 59 43% The first conclusion that can be drawn from Table 2.6 is that even the objective non-language criteria exhibit some instability. This holds in particular for nationality and country of birth. A further analysis of these criteria shows that it is mainly the youngest children who cause the difference. The criterion of self-categorization also turns out to be unstable. In contrast, ethnic identification through home language use and language proficiency is strikingly consistent. Only 1 child and 6 children did not have the same response in the feasibility study and the validation study, respectively. For the other three language criteria, i.e., language choice, language dominance, and language preference, less stable response patterns emerge. Our assumption is that

< previous page

page_30

next page >

< previous page

page_31

next page > Page 31

the number of different responses for language choice was affected by the differences in type of data collection. The results for the Turkish and Moroccan children in grades 1/2 are summarized in Table 2.7. The Moroccan children are divided into Arabic- and Berber-speaking children. Table 2.7 Outcomes of the bilingual receptive vocabulary test for Turkish and Moroccan children, grades 1/2 (* p

< previous page

page_36

next page > Page 36

(4) Language choice:

Do you speak this language with your mother/ father/older brother(s) or sister(s)/younger brother(s) or sister(s)? Do you speak Dutch with your mother/father/older brother(s) or sister(s)/younger brother(s) or sister(s) ?

(5) Language dominance:

Which language do you speak best?

(6) Language preference:

Which language do you like to speak most?

In addition, the teacher asked children in the class for information on language instruction in the pertinent languages: (7) Language need:

Which language(s) would you like to learn at school?

(8) Language participation:

Which language(s) do you learn at school?

The last two questions resulted in a relatively high number of non-responses. Moreover, the phrasing of these questions often did not allow a non-ambiguous interpretation. Therefore, the answers to these questions are not taken into consideration in the present analysis (see Chapter 3, Section 3.2.5, for comparative information on participation in immigrant minority language instruction). The starting point in the selection of children in the language survey was their answer to the questions on country of birth and on home language use. From a total population of 34.451 children, a sample of 9.094 children could be identified as having nonDutch roots. This sample consists of 4.519 boys and 4.337 girls. There were 2.255 children in grades 1/2, 2.259 in grades 3/4, 2.196 in grades 5/6, and 2.146 in grades 7/8; grades 1/2 correspond commonly with ages 4-6 and grades 7/8 commonly with ages 10-12. No information on gender and/or grade was available for the remaining 238 pupils. A more detailed picture of the status and use of specific immigrant minority languages was obtained for a subsample of the research population. The languages focused on in these in-depth studies are Turkish and Arabic (Van der Heijden 1994, Linssen 1994), Sranan and Papiamentu (De Abreu 1994), Khmer (Kipperman 1994), and Romanes (Venes 1994, 1995). Specific questionnaires were administered in interviews with teachers, parents, and other persons involved. In addition to the teacher questionnaire discussed above, questions were asked about language use in the public domain, about interaction patterns in the family, and about needs and attitudes towards the pertinent home languages and towards Dutch. Moreover, for some groups, language vocabulary tests have been used. From the total population of 34.451 children, a sample of 7.179 children (=21%) indicated that in their home another language is used in addition to or instead of Dutch. This sample of 7.179 children of the Home Language Survey was assigned to specific language groups on the basis of geographical area of origin of these languages. Table 2.11 contains an inventory of the home languages

< previous page

page_36

next page >

< previous page

page_37

next page > Page 37

mentioned and the number of times that a particular language was reported by a child to the teacher. The detailed Catalogue of Languages, composed by ILEA (1990), Giacolone Ramat and Ramat (1998), and the Ethnologue: Languages of the World, edited by Grimes (1996), were most helpful in coding and classifying the languages under consideration. Table 2.11 Geographical distribution of reported home languages Asian languages Total Language 1 Language 2 Language 3 Turkish 2082 2073 9 Malay* 250 239 11 Chinese** 277 264 12 1 Hind(ustan)i 198 188 9 1 Vietnamese 54 54 Khmer (Cambodia) 52 52 Javanese 46 39 6 1 Farsi (Iran) 26 26 Kurdish 19 9 10 Pashto (Afghanistan) 17 16 1 Pakistani 16 16 Punjabi/Urdu 14 12 2 Thai 14 14 Bengali (Bangladesh) 12 12 Tagalog (Philippines) 11 10 1 Hebrew 10 8 2 Syrian 7 4 3 Tamil 6 6 Japanese 5 5 Armenian 4 3 1 Gujarati 4 3 1 Lebanese 4 3 1 Hamer (Jemen) 3 3 Iraqi 3 3 Singhalese (Srilanka) 2 1 1 Turoyo 2 2 * As cover term for languages in the Indonesian Archipelago, among which in particular the Moluccan Islands (Javanese is considered to be a separate language). ** As cover term for Cantonese, Hakka, Taiwanese, and Mandarin.

< previous page

page_37

next page >

< previous page

page_38

next page > Page 38

African languages Arabic (North-Africa) Berber 'Moroccan' (Arabic or Berber) Somalian Tigrinese (Eritrea) Swahili (Kenya/Tanzania) Ghanese Dari (Tjaad) Afrikaans (South Africa) Bulu (Cameroon)

Total Language 1 803 647 695 619 633 614

Language 2 149 73 17

Language 3 7 3 2

89 15 6 4 4 2 1

3 2 2 1 1

-

86 13 6 2 4 1 -

European languages

Total Language Language Language 1 2 3 English 411 327 74 10 Spanish (Europe, Latin America) 211 181 27 3 French 134 96 34 4 Serbian/Croatian 106 98 8 German 95 80 14 1 Italian 67 63 3 1 Polish 64 62 2 Portuguese (Portugal, Angola, 42 38 4 Brazil) Greek 18 17 1 Albanese 15 15 Swedish 12 11 1 Danish 10 10 Gaelic (Scotland, Ireland) 9 8 1 Russian 8 8 Czech 8 8 Hungarian 6 6 Roumanian 6 6 Slovenian 5 5 Finnish 3 2 1 Macedonian 3 2 1 Icelandic 3 2 1 Bulgarian 2 2 Slovakian 2 2 Frisian 1 1 Latvian 1 1 Ukrainian 1 1 Other languages Total Language 1 Language 2 Papiamentu 487 465 20 Sranan/Surinamese 381 357 23 Romanes* 128 126 2 Guyanese 3 2 1 Cape Verdian 2 2 * As cover term for Romany, Manish and Sinthi.

< previous page

Language 3 2 1 -

page_38

next page >

< previous page

page_39

next page > Page 39

Unknown Indian Swiss African Belgian Other

Total 21 3 2 1 13

Language 1 17 3 2 1 7

Language 2 4 6

Language 3 -

In the survey, 66 different languages were identified as home languages in addition to or instead of Dutch. In 25 cases, the languages reported function as the only home language and, apart from Dutch, in 41 cases (also) as the second/third home language. The languages that were mentioned more than 20 times as the second/third home language are Arabic (156), English (84), Berber (76), French (38), Spanish (30), Sranan/Surinamese (24), and Papiamentu (22). Eleven languages are presented in Figure 2, i.e., the ten most frequently mentioned languages plus Romanes as the language with the most clearly non-territorial origin. These eleven language groups provided the basis for the analyses and interpretations presented in the next section. Key references which discuss a variety of the most widely used immigrant minority languages in the Netherlands, including the language groups presented in this chapter, are Extra and Verhoeven (1993a, 1993b) Broeder et al (1993), and De Ruiter (1991). Other studies focus on specific (groups of) immigrant minority languages. With respect to Turkish in the Netherlands, studies have been done on the acquisition of Turkish by pre-school or primary school children (Boeschoten 1990, Schaufeli 1991, Aarts 1994, Aarssen 1996), on the vitality of Turkish in secondary education (Özgüzel 1994), and on processes of code-switching between first- and second-generation Turks (Backus 1996). Studies of (Moroccan) Arabic in the Netherlands have focused on its acquisition (Bos 1997), on its status in primary education (Van de Wetering 1990), on code-switching (Nortier 1989, Boumans 1998), and on processes of language maintenance and language loss (El Aissati 1997). Moreover, language proficiency tests have been developed for Turkish and Arabic for use in the initial (Verhoeven et al 1995) and final stages (Aarts and De Ruiter 1995) of primary education, and for the certification at the end of secondary education.

< previous page

page_39

next page >

< previous page

page_40

next page > Page 40

Figure 2 Most frequently mentioned languages in the 1993 Home Language Survey Apart from studies on Turkish and Arabic, few studies have dealt with lesser used immigrant minority languages in the Netherlands. Examples of such studies are Kook (1994) and Narain (1995) on Papiamentu, Charry, Koefoed, and Muysken (1983) and Van der Avoird (1995, 1998) on Surinamese languages, Tahitu (1989) and Rinsampessy (1992) on Moluccan Malay, Tinnemans (1991) and Michielsen (1992) on Italian, Dialektopoulos (1998) on Greek, Research Group Kurds (1996) on Kurdish, and Hendriks (1986) on Chinese. 2.4 Status and Use of the Most Frequently Mentioned Languages A detailed language profile will be given of those children who reported that one of the eleven languages presented in Figure 2 is used at home. First, an overview of the country of birth of the children and their parents is given per language group. Moreover, an overview is presented of the primary language mentioned and other languages that are used at home. The language profile per language group consists of the following four dimensions: Language proficiency With the exception of Berber as an (as yet) uncodified language, the oral and written language proficiency skills reported by the children in each language group will be presented.

< previous page

page_40

next page >

< previous page

page_41

next page > Page 41

Language choice The relative number of children opting for Dutch in interaction with interlocutors at home will be compared with the relative number of children opting for a language other than Dutch. The degree to which a language other than Dutch is always/often spoken with family members provides a good indication of the degree of language maintenance in the immigration context. Language dominance The language(s) which the children say they speak best, i.e., the primary language, Dutch, or the two languages in combination, are taken into account. Language preference For the language(s) the children like to speak most, the procedure is the same as that for specifying language dominance. 2.4.1 Turkish Country of birth The Turkish language group consists of 2082 children. Table 2.12 shows that 85% (1759 children) of this group was born in the Netherlands, whereas most of their parents were born in Turkey. Table 2.12 Country of birth for the Turkish language group Country of birth Pupil Mother Father Netherlands 1759 85% 40 2% 82 4% Turkey 269 13% 1969 95% 1944 93% Belgium 6 4 Germany 4 4 Other/Unknown 44 2% 73 3% 48 3% Total 2082 100% 2082 100% 2082 100% Home language varieties The languages that are used in addition to or instead of Dutch by children in the Turkish language group are given in Table 2.13. For almost all the children, Turkish is the only home language. A small number of pupils reported that Kurdish is used at home. Table 2.13 Languages that are used at home Turkish language group Turkish 2057 Turkish + Kurdish 12 Turkish + Arabic 4 Turkish + other languages 9 Total 2082

< previous page

for the 99% 1% 100%

page_41

next page >

< previous page

page_42

next page > Page 42

Figure 2.1

< previous page

page_42

next page >

< previous page

page_43

next page > Page 43

Language proficiency Figure 2.1 a summarizes the reported language proficiency in Turkish. Almost all of the Turkish pupils say that they speak (2010 children, 98%) and understand (2011 children, 98%) Turkish. A relatively smaller number of children claim reading proficiency (1160 children, 60%) and writing proficiency (1156 children, 59%) in Turkish. Language choice The language choice of the Turkish children in interaction with family members can be derived from Figure 2.1b. A large number of children indicate that they always/often speak Turkish with their father and mother (1.554 children, 78%, and 1.636 children, 80%, respectively). This also holds for the language choice with their younger and older brothers/sisters, but to a lesser degree (593 children, 45%, and 665 children, 58%, respectively). With respect to Dutch, it can be noted that a relatively small number of children always/often speak this language with their mother and father (241 children, 12%, and 285 children, 22%). However, in interaction with younger and older brothers/sisters, Dutch is reported to take a stronger position (629 children, 48%, and 403 children, 35%). Language dominance The language dominance data are summarized in Figure 2.1c. Most of the children in grades 1-4 indicate that Turkish is the language that they speak best (493 children, 88%, and 311 children, 59%). Conversely, most children in grades 5-8 indicate that Dutch is the language they speak best (227 children, 52%, and 270 children, 59%). Language preference The data on language preference are summarized in Figure 2.1d. The pattern that emerges is the same as the one for language dominance of the Turkish children. Whereas most of the younger children prefer to speak Turkish (in grades 1/2, 419 children, 78%, and in grades 3/4, 263 children, 50%), most of the older children prefer to speak Dutch (in grades 5/6, 241 children, 56%, and in grades 7/8, 291 children, 64%). 2.4.2 Arabic Country of birth The Arabic language group consists of 803 pupils, mainly second-generation children. The majority of the parents were born in Morocco (see Table 2.14).

< previous page

page_43

next page >

< previous page

page_44

next page > Page 44

Figure 2.2

< previous page

page_44

next page >

< previous page

page_45

next page > Page 45

Table 2.14 Country of birth for the Country of birth Pupil Netherlands 587 73% Morocco 160 20% Egypt Irak 7 1% Turkey 1 9 Tunisia Algeria Lebanon Syria 2 Somalia 6 1% Other/Unknown 40 5% Total 803 100%

Arabic language group Mother Father 16 2% 34 4% 669 83% 671 84% 17 2% 1 13 2% 10 1% 1% 8 1% 8 1% 1 6 1% 8 1% 5 5 5 5 6 1% 6 1% 49 6% 54 7% 803 100% 803 100%

Home language varieties The home language varieties that are used in addition to or instead of Dutch by children in the Arabic language group are shown in Table 2.15. For 582 children (73%), Arabic is the only home language. For 152 children (19%), in addition to Arabic, Berber is also used at home. Note that the combination of Arabic plus 'Moroccan' may refer to Arabic plus Berber too. Table 2.15 Languages that are used at home Arabic language group Arabic 582 Arabic + Berber 152 Arabic + Moroccan 24 Arabic + French 13 Arabic + Somalian 5 Arabic + English 4 Arabic + Turkish 4 Arabic + Spanish 3 Arabic + other languages 13 Arabic + Berber + Spanish 2 Arabic + Berber + French 1 Total 803

for the 73% 19% 3% 2% 2% 100%

Language proficiency As Figure 2.2a shows, almost all of the 803 pupils in the Arabic language group say that they understand (731 children, 94%) and speak (701 children, 96%) Arabic. With respect to written language proficiency skills, the percentages are much lower: for reading, 40% (298 children), and for writing, 42% (309 children).

< previous page

page_45

next page >

< previous page

page_46

next page > Page 46

Figure 2.3

< previous page

page_46

next page >

< previous page

page_47

next page > Page 47

Language choice The language choice pattern within the families is given in Figure 2.2b. Most children say that they always/often speak Arabic with their parents: 459 children (60%) with the mother, and 445 children (60%) with the father. A mirror-like picture appears for the use of Dutch. With their younger as well as their older brothers/sisters, most children always/often speak Dutch (306 children, 66%, and 418 children, 54%, respectively). Language dominance The language dominance pattern is presented in Figure 2.2c. The children in the youngest grades say that Arabic is the language they speak best (in grades 1/2, 90 children, 61%). On the other hand, the older pupils (grades 3/4, 5/6 and 7/8) indicate that their best language is Dutch (106 children, 52%; 102 children, 62%, and 114 children, 70%, respectively). Language preference Preference for Arabic (see Figure 2.2d) is also found at a younger age (grades 1/2, 71 children, 52%; grades 3/4, 102 children, 51%). At a later age (grades 5/6, 99 children, 57%; grades 7/8, 103 children, 62%), a shift in preference towards Dutch can be noted. In this respect, the language preference pattern seems to be a reflection of the language dominance pattern. 2.4.3 Berber Country of birth Most of the 695 Berbers belong to the second generation of Moroccans in the Netherlands (see Table 2.16). Most of the parents were born in Morocco (667 children, 96%; 655 children, 94%). Table 2.16 Country of birth for the Berber language group Country of birth Pupil Mother Father Netherlands 490 71% 4 1% 7 1% Morocco 191 27% 667 96% 655 94% Other/Unknown 14 2% 24 3% 33 5% Total 695 100% 695 100% 695 100% Home language varieties Table 2.17 gives an overview of the language varieties used at home by the children in the Berber language group. There is a relatively high number of 498 children who speak Berber at home in addition to or instead of Dutch (72%). 152 children (22%) also speak Arabic. The combination of Berber plus 'Moroccan' might also refer to Berber plus Arabic.

< previous page

page_47

next page >

< previous page

page_48

next page > Page 48

Table 2.17 Languages that are used at home Berber language group Berber 498 Berber + Arabic 152 Berber + Moroccan 38 Berber + French 4 Berber + English 1 Berber + Arabic + French 1 Berber + Arabic + Spanish 1 Total 695

for the 72% 22% 6% 100%

Language proficiency Almost all the children responded that they can understand (660 children, 96%) and speak (645 children, 94%) a Berber language. Reading and writing skills in Berber were not taken into account. Language choice Most Berber children always/often speak Berber with their father and mother (517 children, 76%, and 447 children, 68%, respectively). However, they generally choose Dutch in interaction with their younger and older brothers and sisters (318 children, 60%, and 399 children, 68%, respectively). Language dominance Figure 2.3c shows which language these children report they speak best. At the start of their primary school careers, most first and second graders say that Berber is their best language (89 children, 76%). However, for older age groups (grades 7/8), a shift towards Dutch emerges. Most of the children report Dutch as their dominant language at the end of their primary school education (143 children, 71%). Language preference Language preference (see Figure 2.3d) also shows a pattern of shift. Most Berber children in the lower grades prefer to speak Berber (85 children, 71%), whereas the children in the higher grades prefer Dutch (132 children, 67%). 2.4.4 Papiamentu Country of birth Almost half of the 487 Papiamentu children were born in the Netherlands, whereas the majority of their parents were born in the Dutch Antilles.

< previous page

page_48

next page >

< previous page

page_49

next page > Page 49

Table 2.18 Country of birth for the Papiamentu language group Country of birth Pupil Mother Father Netherlands 226 46% 26 5% 42 9% Antilles 242 50% 376 77% 387 79% Surinam 2 31 7% 21 4% South/Central America 3 1% 5 1% 5 1% Other/Unknown 14 3% 49 10% 32 7% Total 487 100% 487 100% 487 100% Home language varieties Table 2.19 lists the languages that are used at home in addition to or instead of Dutch by the Antillean children. In most families, only Papiamentu is spoken. 51 children (11%) also speak Spanish and/or English in addition to Papiamentu. Table 2.19 Languages that are used at home for the Papiamentu language group Papiamentu 425 87% Papiamentu + English 29 6% Papiamentu + Spanish 20 4% Papiamentu + Spanish + English 2 1% Papiamentu + other languages 11 2% Total 487 100% Language proficiency Almost all children in the Papiamentu language group say that they understand (430 children, 91%) and speak (388 children, 83%) Papiamentu. Reading and writing skills in Papiamentu are less common (147 children, 32%, and 117 children, 26%, respectively). Language choice In spite of their oral proficiency in Papiamentu, Dutch is used prominently in interaction with family members (see Figure 2.4b): 264 children (55%) always/often speak Dutch with their mother, 255 children (55%) with their father, 173 children (61%) with their younger brothers and sisters, and 177 children (65%) with their older brothers and sisters. Language dominance According to the language dominance profile, most children in the Papiamentu group say that Dutch is the language they speak best. This observation can be made for all grades (see Figure 2.4c): in grades 1/2, 62 children (64%), in grades 3/4, 70 children (65%), in grades 5/6, 91 children (72%), and in grades 7/8, 83 children (68%).

< previous page

page_49

next page >

< previous page

page_50

next page > Page 50

Figure 2.4

< previous page

page_50

next page >

< previous page

page_51

next page > Page 51

Language preference Dutch is also the language that the majority of children in all grades prefer to speak (see Figure 2.4d): in grades 1/2, 59 children (63%), in grades 3/4, 64 children (61%), in grades 5/6, 81 children (64%), and in grades 7/8, 68 children (58%). 2.4.5 English Country of birth Table 2.20 gives an overview of the countries of birth for the English language group. Most of the children and their parents were born in the Netherlands. Apart from Great Britain, Australia, New Zealand, and the United States, a relatively large number of families for whom English is the home language, come from the Dutch Antilles. Table 2.20 Country of birth for the English Country of birth Pupil Netherlands 252 61% Great Britain 25 6% Antilles 16 2% Australia/New Zealand 11 3% South Africa 10 2% United States 10 2% Surinam 4 1% Canada 3 1% Ireland 1 1% Africa 17 4% Asia 19 5% South/Central America 18 4% Europe 12 3% Other/Unknown 13 3% Total 411 100%

language group Mother Father 128 31% 151 37% 100 24% 56 14% 40 10% 42 10% 7 2% 10 2% 7 2% 10 2% 9 2% 7 2% 9 2% 11 3% 4 1% 4 1% 5 1% 28 7% 26 6% 27 7% 35 8% 22 5% 23 6% 6 2% 23 6% 19 5% 13 3% 411 100% 411 100%

Home language varieties The language varieties that are used at home for the English language group are given in Table 2.21. In most homes, only English is used instead of or in addition to Dutch (279 children, 68%). Other languages used in this group of children are Papiamentu, Spanish, and languages of ex-colonial countries. English functions mostly as lingua franca in these countries.

< previous page

page_51

next page >

< previous page

page_52

next page > Page 52

Figure 2.5

< previous page

page_52

next page >

< previous page

page_53

next page > Page 53

Table 2.21 Languages that are used at home for the English language group English 279 68% English + Spanish 10 2% English + Papiamentu 34 8% English + African languages 22 5% English + Asian languages 28 7% English + European languages 28 7% English + other languages 10 2% Total 411 100% Language proficiency As Figure 2.5a shows, most children indicate that they understand (343 children, 90%) and speak (319 children, 84%) English. Fewer children indicate that they read (162 children, 45%) and write (115 children, 32%) English. Language choice At home, most children always/often speak Dutch with their family members. This figure is higher in interaction with younger and older brothers/sisters (150 children, 74%, and 182 children, 76%, respectively) than with their mothers (250 children, 62%) and fathers (228 children, 61%). Language dominance The reported proficiency of the English language group corresponds with the reported language dominance. Most children in all grades increasingly indicate that their best language is Dutch: in grades 1/2, 45 children (68%), in grades 3/4, 65 children (74%), in grades 5/6, 89 children (79%), and in grades 7/8, 79 children (80%). Language preference Most of the children in all grades also prefer speaking Dutch: in grades 1/2, 45 children (69%), in grades 3/4, 48 children (56%), in grades 5/6, 64 children (58%), and in grades 7/8, 47 children (55%). From grade 3 onwards, however, the number of children who claim that English is their favoured language is higher than the number of children who say that English is their best language (compare Figures 2.5c and 2.5d). 2.4.6 Sranan/Surinamese. The Surinamese language group consists of 381 children. Firstly, this includes the 80 children who indicate that Sranan (Tongo) is spoken at home ('Sranan Tongo' literally means 'Surinamese language'). The remaining 301 children say that 'Surinamese' is used in their family in addition to Dutch. These children do not specify further which Surinamese language variety is spoken.

< previous page

page_53

next page >

< previous page

page_54

next page > Page 54

Figure 2.6

< previous page

page_54

next page >

< previous page

page_55

next page > Page 55

Country of birth Most of the children in this language group (298 children, 78%) belong to the second generation born in the Netherlands. In contrast, their parents were predominantly born in Surinam: 295 mothers (77%) and 328 fathers (86%). Table 2.22 Country of birth for the Sranan/Surinamese language group Country of birth Pupil Mother Father Netherlands 298 78% 33 9% 34 9% Surinam 67 18% 295 77% 328 86% Antilles 1 6 2% 3 1% Other/Unknown 15 4% 47 12% 16 4% Total 381 100% 381 100% 381 100% Home language varieties The languages that are used in addition to or instead of Dutch by children in the Surinamese language group are given in Table 2.23. 266 children (70%) say that 'Surinamese' is spoken at home. They are probably referring to Sranan. However, this is not certain. In fact, these children may have not been able to provide a precise specification of the home language. In addition to Sranan/Surinamese, for some children one of the following languages is used at home: Hind(ustan)i, Javanese, English, and Papiamentu. Table 2.23 Languages that are used at home for the Sranan/Surinamese language group Surinamese 266 Sranan 69 Surinamese + Sranan 2 Surinamese + Hind(ustan)i 9 Sranan + Hind(ustan)i 7 Surinamese + Javanese 7 Sranan + Javanese 4 Surinamese + English 8 Surinamese + Papiamentu 4 Surinamese + other languages 5 Total 381

70% 18% 1% 2% 2% 2% 1% 2% 1% 1% 100%

Language proficiency A large number of Surinamese children indicated that they understand Sranan/Surinamese (294 children, 81%). This figure is considerably lower for speaking proficiency (211 children, 63%), and still lower for reading (103 children, 29%), and writing proficiency (61 children, 17%) in Sranan/Surinamese (Figure 2.6a).

< previous page

page_55

next page >

< previous page

page_56

next page > Page 56

Figure 2.7

< previous page

page_56

next page >

< previous page

page_57

next page > Page 57

Language choice Given the relatively large number of children who report that they understand Sranan/Surinamese, the language choice profile of the Surinamese children is remarkable. Figure 2.6b shows that predominantly Dutch is spoken with the mother (296 children, 80%), father (251 children, 78%), younger brothers/sisters (181 children, 85%), and older brothers/sisters (233 children, 85%). Language dominance As can be derived from Figure 2.6c, most of the Surinamese children in grades 18 indicated that their best language is Dutch: in grades 1/2, 30 children (70%), in grades 3/4, 71 children (82%), in grades 5/6, 106 children (91%), and in grades 7/8, 104 children (92%). Language preference For all grades, most of the children say the language they like most is also Dutch: in grades 1/2, 26 children (63%), in grades 3/4, 61 children (73%), in grades 5/6, 71 children (65%), and in grades 7/8, 86 children (77%). Nevertheless, a comparison of Figures 2.6c and 2.6d also reveals that for all grades, the number of children with a preference for Sranan/Surinamese exceeds the number of children with a dominance in Dutch. 2.4.7 Chinese Country of birth The majority of the children in the Chinese language group were born in the Netherlands (219 children, 79%). Most of their parents come from the Chinese Republic or from Hong Kong. Table 2.24 Country of birth for the Country of birth Pupil Netherlands 219 79% Chinese Republic 21 8% Great Britain 1 Surinam 3 1% Vietnam 1 Indonesia Hong Kong 11 4% Malaysia 4 1% Singapore Taiwan 2 1% Other/Unknown 15 6% Total 277 100%

< previous page

Chinese language group Mother Father 4 1% 2 1% 125 45% 130 47% 7 3% 6 2% 2 1% 1 7 3% 5 2% 7 3% 5 2% 78 28% 79 29% 17 6% 17 6% 6 2% 4 1% 5 2% 8 3% 19 6% 20 7% 277 100% 277 100%

page_57

next page >

< previous page

page_58

next page > Page 58

Home language varieties Most of the children live in a home context in which exclusively Chinese is spoken in addition to or instead of Dutch (266 children, 96%). A few children are also exposed to Malay, English, or Vietnamese. Table 2.25 Languages that are used at home for the Chinese language group Chinese 266 96% Chinese + Malay 5 2% Chinese + English 4 2% Chinese + Vietnamese 1 Chinese + Malay + English 1 Total 277 100% Language proficiency The reported language proficiency of the Chinese children is presented in Figure 2.7a. The number of children who indicate that they understand (264 children, 98%) and speak (258 children, 95%) Chinese is remarkably high. Reading and writing proficiency are much less developed (75 children, 31%, and 82 children, 34% respectively). Language choice Most of the children indicated that they always/often speak Chinese with their mother (206 children, 77%) and father (206 children, 78%). In interaction with brothers and sisters, Dutch has a more prominent position: 56 children (39%) always/often speak Chinese with younger brothers/sisters, 66 children (35%) always/often with older brothers/sisters. Language dominance From the language dominance profile (Figure 2.7c), it becomes clear that in particular in the early grades 1-4, most of the children feel that their best language is Chinese: in grades 1/2, 39 children (74%), and in grades 3/4, 28 children (46%). In the later grades 5-8, most of the children refer to Dutch as their dominant language: in grades 5/6, 35 children (51%), and in grades 7/8, 53 children (66%). Language preference The language preference profile (Figure 2.7d) shows a similar language shift pattern from Chinese towards Dutch during the primary school career. In grades 1/2, Chinese is the language the children like to speak most (32 children, 68%). In later grades, the preferred language children report is Dutch: in grades 3/4, 34 children (60%), in grades 5/6, 42 children (61%), and in grades 7/8, 59 children (73%).

< previous page

page_58

next page >

< previous page

page_59

next page > Page 59

2.4.8 Malay Country of birth Of the 250 children in the Malay language group, 91% (227 children) were born in the Netherlands. Of their parents, about one third were born in the Netherlands and about one-half are of Moluccan/Indonesian origin. Table 2.26 Country Country of birth Netherlands Indonesia Surinam Malaysia Other/Unknown Total

of birth for the Malay language group Pupil Mother Father 227 91% 83 33% 95 38% 12 5% 134 54% 124 50% 3 1% 1 2 1% 5 2% 7 3% 9 3% 25 10% 23 9% 250 100% 250 100% 250 100%

Home language varieties (Moluccan) Malay is the language that is mostly used in addition to or instead of Dutch. Moreover, in some families, Chinese and English are also used as home languages (Table 2.27). Table 2.27 Languages that are used at home Malay language group Malay 233 Malay + Chinese 5 Malay + English 5 Malay + other languages 7 Total 250

for the 93% 2% 2% 3% 100%

Language proficiency The language proficiency profile of the Malay language group shows that Malay is understood by 68% (159 children) of the children and spoken by 58% (135 children). These figures are remarkably high, given the pattern of language choice in these families. Reading and writing proficiency in Malay is reported by far fewer children (53 children, 23%, and 38 children, 16% respectively). Language choice For a small number of the children, Malay is always/often spoken with the parents (30 children, 13%, and 27 children, 12% respectively). For an even smaller number, Malay is spoken with older and younger brothers/sisters (10 children, 6% and 5 children, 4% respectively). By far the most children always/often speak Dutch in interaction with other family members.

< previous page

page_59

next page >

< previous page

page_60

next page > Page 60

Figure 2.8

< previous page

page_60

next page >

< previous page

page_61

next page > Page 61

Language dominance The observations for language dominance are summarized in Figure 2.8c. In all grades, most of the children say that the language they speak best is Dutch: in grades 1/2, 34 children (85%), in grades 3/4, 56 children (95%), in grades 5/6, 70 children (96%), and in grades 7/8, 64 children (92%). Language preference The reported language preference of these children in all grades also points towards a strong preference for Dutch: in grades 1/2, 33 children (82%), in grades 3/4, 46 children (81%), in grades 5/6, 47 children (66%), and in grades 7/8, 48 children (71%). From grades 1/2 onwards, however, Malay is mentioned relatively more often as the preferred language than as the dominant language. 2.4.9 Spanish Country of birth For 211 children, Spanish is spoken at home. Table 2.28 gives an overview of the countries in which the children of this language group and their parents were born. Most of the Spanish-speaking children were born in the Netherlands (140 children, 66%). Most of their parents were born outside the Netherlands, most commonly in Spain. The latter holds for 83 mothers (39%) and 79 fathers (37%). Table 2.28 Country of birth for the Spanish language group Country of birth Pupil Mother Father Netherlands 140 66% 53 25% 40 19% Spain 21 10% 83 39% 79 37% Antilles 11 5% 28 13% 23 11% South/Central America 20 10% 21 10% 44 21% Other/Unknown 19 9% 26 13% 25 12% Total 211 100% 211 100% 211 100% Home language varieties Table 2.29 shows which languages are used at home for the Spanish group. In most families, Spanish is used exclusively in addition to or instead of Dutch (169 children, 80%). Moreover, in a number of families Papiamentu and/or English are also spoken.

< previous page

page_61

next page >

< previous page

page_62

next page > Page 62

Figure 2.9

< previous page

page_62

next page >

< previous page

page_63

next page > Page 63

Table 2.29 Languages that are used at home for the Spanish language group Spanish 169 80% Spanish + Papiamentu 14 7% Spanish + English 10 5% Spanish + Papiamentu + English 5 2% Spanish + Arabic 4 2% Spanish + other languages 9 4% Total 211 100% Language proficiency Figure 2.9a gives the results for the four language skills. A relatively high number of children in the Spanish group reported that they can understand (182 children, 89%) and speak (179 children, 85%) Spanish. Reading and writing skills are reported by smaller numbers of children (90 children, 46% for reading, and 81 children, 42% for writing). Language choice Spanish is used by most of the children with their parents as the language of interaction: 93 children (46%) always/often speak Spanish with their mother and 89 children (47%) always/often with their father. A change can be noted for the pattern of language choice in interaction with younger and older brothers/sisters. The number of children opting for Dutch in these interactions is substantially higher: 57 children (66%) always/often speak Dutch with younger brothers/sisters and 97 children (78%) with older brothers/sisters. Language dominance The language dominance profile for the children in the Spanish group shows that for all grades, most children say that the language they speak best is Dutch: in grades 1/2, 19 children (61%), in grades 3/4, 36 children (88%), in grades 5/6, 35 children (78%), and in grades 7/8, 43 children (72%). Language preference The preferred language is also Dutch for all grades: in grades 1/2, 19 children (63%) refer to Dutch as the preferred language, in grades 3/4, 27 children (68%), in grades 5/6, 27 children (59%), and in grades 7/8, 33 children (59%). Nevertheless, for all grades, the number of children with a preference for Spanish is somewhat higher than the number of children who report dominance in Dutch. 2.4.10 Hind(ustan)i Country of birth The Hind(ustan)i language group consists of 198 children. As can be seen in Table 2.30, these are mainly SurinameseHind(ustan)i children who were born in the Netherlands (178 children, 90%) and whose parents were born in Surinam.

< previous page

page_63

next page >

< previous page

page_64

next page > Page 64

Figure 2.10

< previous page

page_64

next page >

< previous page

page_65

next page > Page 65

Table 2.30 Country group Country of birth Netherlands Surinam India Pakistan Other/Unknown Total

of birth for the Hind(ustan)i language Pupil 178 90% 14 7% 1 5 3% 198 100%

Mother 10 5% 160 80% 7 4% 2 1% 19 10% 198 100%

Father 2 1% 183 92% 1 12 7% 198 100%

Home language varieties For the families in question, Hind(ustan)i is generally the only language used in addition to or instead of Dutch (177 children, 89%). Sranan/Surinamese is reported as a third language for only a few families (11 children, 5%). Table 2.31 Languages that are used at home for the Hind(ustan)i language group Hind(ustan)i 177 89% Hind(ustan)i + Sranan/Surinamese 11 5% Hind(ustan)i + English 4 2% Hind(ustan)i + Urdu/'Indian' 3 2% Hind(ustan)i + other languages 3 2% Total 198 100% Language proficiency Most children understand Hind(ustan)i (162 children, 89%). A somewhat smaller number of children also speak Hind(ustan)i (123 children, 67%). The number of children who report that they can read and write Hind(ustan)i (19 children, 10%, and 16 children, 9%, respectively) is remarkably low (Figure 2.10a). Language choice Although a substantial number of children report oral language proficiency skills in Hind(ustan)i, in most families, Dutch is the most important language for interaction with all family members (Figure 2.10b): 141 children (73%) always/often speak Dutch with their mother, 128 children (71%) with their father, 98 children (82%) with younger brothers/sisters, and 105 children (81%) with older brothers/sisters. Language dominance A large number of children in the Hind(ustan)i language group say that their best language is Dutch. This holds for all grades (Figure 2.10c): in grades 1/2, 28 children (80%), in grades 3/4, 34 children (63%), in grades 5/6, 40 children (80%), and in grades 7/8, 48 children (87%).

< previous page

page_65

next page >

< previous page

page_66

next page > Page 66

Figure 2.11

< previous page

page_66

next page >

< previous page

page_67

next page > Page 67

Language preference The language these children say they like to speak most appears to be most often Dutch: in grades 1/2, 25 children (76%), in grades 3/4, 32 children (60%), in grades 5/6, 30 children (64%), and in grades 7/8, 43 children (78%). 2.4.11 Romanes Country of birth 'Romanes' is used as the cover term for Romany, Manish, and Sinthi. The Romanes language group consists of 128 children. The majority of the children and their parents were born in the Netherlands (see Table 2.32). Table 2.32 Country of birth for the Romanes language group Country of birth Pupil Mother Father Netherlands 121 95% 99 78% 90 70% Belgium 3 1% 7 5% France 5 4% 2 2% Morocco 1 1% 3 2% 3 2% Italy 2 1% Germany 1 1% 1 1% Sweden 1 1% 1 1% Former Yugoslavia 9 7% Norway 2 2% Other/Unknown 4 3% 16 13% 13 10% Total 128 100% 128 100% 128 100% Home language varieties The home language varieties that are used in addition to or instead of Dutch by the children in the Romanes language group are shown in Table 2.33. Romanes clearly has a strong position. For 125 children (99%), Romanes is the only language at home. Only 3 children (1%) report that in addition to Romanes German or 'Moroccan' is also used at home; 8 children label their home language as 'Zigeunertaal' (Gypsy language). Table 2.33 Languages that are used at home Romanes language group Romany 50 Manish 15 Sinthi 52 Gypsy language 8 Romanes + Moroccan 2 Romanes + German 1 Total 128

< previous page

for the 39% 12% 41% 7% 1% 100%

page_67

next page >

< previous page

page_68

next page > Page 68

Language proficiency The language proficiency profile of the Romanes language group shows that Romanes is understood by 99% (127 children) of the children and spoken by 97% (124 children). These figures are remarkably high. Reading and writing proficiency in Romanes are reported by few children (20 children, 18% and 21 children, 19% respectively). Language choice The language choice pattern is presented in Figure 2.1 lb. A high number of children reported using Romanes always/often with family members: 107 children (85%) with the mother, 106 children (86%) with the father, 84 children (87%) with older brother/sisters, and 83 children (89%) with younger brothers/sisters. A mirror-like picture appears for the use of Dutch. Only few children in the Romanes language group report that they always/often speak Dutch with their mother (15 children, 12%), with their father (13 children, 11%), with their older brothers/sisters (14 children, 14%), or with their younger brothers/sisters (13 children, 14%). Language dominance The observations for language dominance are summarized in Figure 2.1 c. In all grades, most of the children say that the language they speak best is Romanes: in grades 1/2, 26 children (88%), in grades 3/4, 32 children (88%), in grades 5/6, 20 children (62%) and, in grades 7/8, 21 children (92%). Language preference The language preference of these children in all grades also points towards a strong preference for Romanes: in grades 1/2, 27 children (90%), in grades 3/4, 28 children (82%), in grades 5/6, 18 children (56%), and in grades 7/8, 20 children (87%). 2.5 Conclusions and Discussion On the basis of the profiles that have been specified for the eleven language groups under consideration, a cross-linguistic comparison was made which reveals the relative position of each language group. With respect to country of birth, Table 2.34 presents, in decreasing order, the relative number of children per language group that were born in the Netherlands. For comparison, the number of parents born in the Netherlands is also given.

< previous page

page_68

next page >

< previous page

page_69

next page > Page 69

Table 2.34 Order of children born in the Netherlands (in %) Language group Child Mother Father N total Romanes 95 78 70 128 Malay 91 33 38 250 Hind(ustan)i 90 5 1 198 Turkish 85 2 4 2082 Chinese 79 1 1 277 Sranan/Surinamese 78 9 10 381 Arabic 73 2 4 803 Berber 71 1 695 Spanish 66 25 19 211 English 61 31 37 411 Papiamentu 46 5 9 487 With the exception of the Papiamentu language group, most of the children within each language group were born in the Netherlands. Moreover, most of the parents in all language groups were born outside the Netherlands. The children in the Malay language group (predominantly of Moluccan origin) and even more the children in the Romanes language group generally belong to the third generation of immigrant minority groups in the Netherlands. Table 2.35 presents a comparison of all language groups on the basis of the following five dimensions: • language competition: the language does not compete with other languages, apart from Dutch; • language proficiency: the language is understood by the children; • language choice: the language is often/always used with the mother; • language dominance: the language is spoken best by the youngest children (grades 1/2); • language preference: the language is liked most by the youngest children (grades 1/2). In the final column of Table 2.35, these five language profile dimensions are summarized in a (decreasing) language vitality index, based on the average value of the percentaged scores for each of the five dimensions. The resulting language vitality index is obviously arbitrary in the sense that different dimensions are equally weighted.

< previous page

page_69

next page >

< previous page

page_70

next page > Page 70

Table 2.35 Language vitality index per language group, derived from five language profile dimensions (in %) Language Language Language Language Language Language Language group competition proficiency choice dominance preference vitality Romanes 99 99 85 88 90 92.2 Turkish 99 98 80 88 78 88.6 Chinese 96 98 77 74 68 82.6 Berber 72 96 76 76 71 78.2 Arabic 73 94 60 61 52 68.0 Papiamentu 87 91 41 33 35 57.4 Spanish 80 89 46 32 30 55.4 English 68 90 33 28 29 49.6 Sranan/Surin. 88 81 15 26 32 48.4 Hind(ustan)i 89 89 24 14 18 46.8 Malay 93 68 13 15 18 41.4 Table 2.35 makes clear that there is strong variation in the vitality of immigrant minority languages in the Netherlands. A comparison of the ranking on the basis of country of birth (see Table 2.34) and the ranking based on the language vitality index (see Table 2.35) reveals a low correlation, indicating that the degree of vitality of a language group does not necessarily correlate strongly with the degree to which the children of the pertinent language group are born in the Netherlands. The strongest vitality can even be found for the Romanes language group which consists mostly of third-generation children. None of the home languages are in strong competition with languages other than Dutch. Arabic and Berber co-occur as home languages relatively often within Moroccan families. English is frequently a lingua franca in a context in which other languages are also used at home. With respect to language proficiency, it can be noted that reading and writing skills are reported to be relatively weak compared to oral skills in the home languages. Literacy is a skill that is typically acquired and enhanced in a school context which for most of the languages under discussion is rarely the case in the Netherlands. The highest percentages of children who can read in their home language were found for Turkish (60%), Spanish (46%), English (45%), and Arabic (40%). Scores below 40% emerge for Papiamentu (32%), Chinese (31%), Sranan/Surinamese (29%), Malay (23%), Romanes (18%), and Hind(ustan)i (10%). Berber was not been taken into account in this regard. The pattern of majority language choice (Dutch) of children in interaction with family members is presented in Table 2.36 (again in percentages).

< previous page

page_70

next page >

< previous page

page_71

next page > Page 71

Table 2.36 Choice of Dutch by children in interaction with family members (in %) Language group MotherFather Younger Older siblings siblings Romanes 12 11 14 14 Turkish 12 22 35 48 Chinese 11 13 52 39 Berber 10 20 60 68 Arabic 18 21 66 54 Papiamentu 55 55 61 65 Spanish 44 46 66 78 English 62 61 74 76 Sranan/Surinamese 80 78 85 85 Hind(ustan)i 73 71 82 81 Malay 82 85 92 92 The data presented in Table 2.36 are a mirror-like confirmation of a classic pattern of language shift. The mother emerges as the gatekeeper of language maintenance (thus supporting the widely used concept of 'mother tongue'), whereas the father takes a strong second position. Shift of choice towards the majority language (Dutch) increasingly runs via younger and older siblings, respectively. Notable exceptions in this language choice pattern are the Chinese and Arabic language groups. Again, the Romanes language group takes a special position. The data on language dominance and language preference in the preceding sections have been presented for different grades (Gl/2, G3/4, G5/6, G7/8). These data exhibit a cross-sectional picture of development over age. With respect to language dominance, 6 out of 11 language groups (Turkish, Arabic, Berber, English, Sranan/Surinamese, and Chinese) show a scissor-like pattern in which dominance in the home language decreases with age, and in which dominance in Dutch increases with age. For two language groups, a slight recovery of the dominance in the home language emerges in grades 7/8 (Papiamentu and Malay). For the Spanish language group, such recovery already starts in grades 5/6. It should be noted again that these cross-sectional data are in need of longitudinal confirmation. With respect to language preference, a similar scissor-like pattern (i.e., decreasing preference for the minority language and increasing preference for the majority language Dutch) emerges for 4 out of 11 language groups (Turkish, Arabic, Berber, and Chinese). For 4 other language groups, preference for the minority language in the older age groups (grades 7/8) is higher than in the youngest age groups (grades 1/2) (Papiamentu, English, Malay, and Spanish). In Table 2.37, the dimensions of language dominance and language preference are featured. For this analysis, children were selected whose answers to the questions on language dominance and language preference show a pattern of shift. Those children who opted for both languages were considered to belong to the groups with dominance/preference in the minority language.

< previous page

page_71

next page >

< previous page

page_72

next page > Page 72

Table 2.37 Absolute and relative number of mismatches between language preference and language dominance Language Number Preferent in Preferent in Total group of Dutch, other children dominant language, in other dominant in language Dutch Turkish 1988 322 16% 179 9% 50125% Arabic 770 120 16% 126 16% 24632% Berber 667 87 13% 122 18% 20931% Papiamentu 468 54 11% 88 19% 14230% English 394 33 8% 92 23% 12531% Sranan/Surin.366 16 4% 74 20% 90 24% Chinese 263 50 19% 28 11% 78 30% Malay 242 2 51 21% 53 21% Spanish 193 19 10% 44 23% 63 33% Hind(ustan)i 192 6 3% 23 12% 29 15% Romanes 128 8 6% 4 3% 12 9% For a relatively high number of children, the language they like most is the language they do not speak best. Table 2.37 indicates an unequivocal difference, in particular for the Turkish and Chinese language groups. Although most children in these groups report that their best language is Turkish/Chinese, the language they like to speak most is Dutch. In contrast, many children in the English, Surinamese, Malay, and Spanish groups express a preference for their heritage language, whereas their best language is Dutch. With respect to Sranan, Malay, and Spanish, this shift may be caused by their status as immigrant minority languages that are relatively infrequently spoken. The smallest mismatch between language dominance and language preference emerges for the Romanes language group. As Table 2.35 made clear, Romanes is also the language with the highest vitality index. The outcomes of home language surveys like the present one are relevant from at least three different perspectives. First of all, home language surveys offer promising evidence on the value of the home language criterion for the identification of multicultural population groups in (school) statistics (see also Section 2.2.3). Secondly, home language surveys offer a wealth of comparative cross-linguistic data on the vitality of different immigrant minority languages, in particular immigrant minority languages for which little other empirical evidence is available (see also the references given at the end of Section 2.3). Thirdly, home language survey data offer indispensable corner stones for educational policy on immigrant minority children, in particular in the domain of immigrant minority language instruction. The last mentioned perspective will be taken up in the next chapter.

< previous page

page_72

next page >

< previous page

page_73

next page > Page 73

3 Educational perspectives: Case studies on the status of immigrant minority languages in education. 3.1. Aims and Method In this chapter, we examine the policies of a number of European Union countries regarding immigrant minority language instruction in both primary and secondary education (see also Broeder and Extra 1997b, 1997c). The cross-national terminology for this type of instruction is not consistent, as can be derived from designations like home language instruction, instruction in the native language and culture, instruction in immigrant languages, or instruction in ethnic minority languages. We will use the acronym IMLI (Immigrant Minority Language Instruction) when referring to this type of instruction in the countries under consideration. The decision to use the designation IMLI is motivated by the inclusion of a broad spectrum of potential target groups. As has been made clear in Chapter 2, the status of an immigrant minority language as the 'native' or home language can change through intergenerational processes of language shifts. Moreover, in secondary education, both minority and majority pupils are often de jure (although seldom defacto) admitted to IMLI (in the Netherlands, e.g., Turkish is a secondary school subject called 'Turkish' rather than 'home language instruction'; see also the concept of Enseignement des Langues et Cultures d'Origine versus Enseignement des Langues Vivantes in the section on France). The extent to which IMLI is a topic in government policy with European dimensions is related to estimates of the number of foreign resident groups in the countries of the European Union presented by the Statistical Office of the EU (EuroStat) in Luxembourg (see Chapter 1, Section 1.2.1). These estimates are based on the conservative nationality criterion. The largest groups, each with more than two million speakers, are the Turkish and Maghreb communities (members of the Western Maghreb community originate from Morocco, Algeria, or Tunisia). Within the EU, the majority of Turks live in Germany, and the majority of Moroccans, in France. The Netherlands is the EU country where the second largest groups of Turks and Moroccans have settled. As a result of processes of migration and minorization, immigrant minority languages like Turkish and Arabic have acquired greater significance in the European context, when measured by the size of the respective groups, than have indigenous or regional minority languages like Basque, Welsh, or Frisian. For various reasons, the development of an educational policy regarding IMLI was, and continues to be, a complex and challenging task. In view of the multicultural composition of many schools, this task involves the organization of multilingual rather than bilingual instruction. Experiences with, and the results of research into, an exclusively bilingual context are therefore only transferable to a

< previous page

page_73

next page >

< previous page

page_74

next page > Page 74

limited degree. In addition, there are big differences as to the nature and extent of bilingualism of immigrant minority pupils, both within and across different language groups (see also Chapter 2). Moreover, from an intergenerational perspective, these differences tend to increase and shift in the direction of the dominant language of the immigrant country. Furthermore, given the very divergent target groups, it is no easy task to fit IMLI into the rest of the curriculum. In a number of countries, the current policy is ambivalent in the sense that, for some groups, IMLI is an addition to the curriculum, while for others, IMLI replaces a part of the curriculum. Finally, there is the question of feasibility in the case of a relatively modest demand for instruction and of relatively small or widely scattered groups. The cross-national comparison of the countries in this chapter is based on secondary analyses of the available data and on oral or written information supplied by key informants (see Appendix 2 for the utilized questionnaire). For each country, there will be a context-bound introduction, followed by a discussion of three key dimensions of IMLI, namely, arguments and objectives, target groups and enrolment, and implementation and organization. 3.2. Educational Experiences in European Union Countries The focus is on three EU countries with relatively large numbers of immigrant minority groups (Germany, France, Great Britain), on two countries which partially share their language of public use (The Netherlands and Flanders/Belgium) and on one of the Scandinavian countries (Sweden). The presentation of the outcomes on each of these six countries will be in alphabetical order. 3.2.1 Belgium1 Belgium is traditionally referred to as a bilingual country with historically determined and comprehensive legislation with regard to the status and use of Dutch and French. Since 1987, Belgium has been a federal state in which the responsibility and authority with respect to migration and education policy have been shifted to the Flemish and Walloon regional governments. In this section, the emphasis will be on IMLI in Flanders and on the status of IMLI in relation to that of instruction in Dutch. Damen (1998) focuses on instruction in Arabic for Moroccan children in Flanders. 1 The section on IMLI in Belgium/Flanders is mainly based on information of the Department of Education and the Flemish Educational Council (VLOR). Special thanks are due to Fernand Snytsers (Brussels) and Suzanne Unck (Tilburg).

< previous page

page_74

next page >

< previous page

page_75

next page > Page 75

Arguments and objectives A major educational goal with regard to children of immigrant minority groups in Flanders is to redress their disadvantaged position, and, in this framework, there is a strong emphasis on learning Dutch. From this perspective, IMLI plays only a peripheral role. The available governmental facilities are deployed for instruction in Dutch, intercultural education, the prevention and solution of developmental or learning problems, and school community work (cf. Omzendbrief 'circular' 1996). IMLI has been officially available since 1981. The original aim of this type of instruction was to prepare the children of parents from specific labour force recruitment countries for remigration. While this goal has now been almost completely abandoned, it has not been replaced by a clear new aim. As a consequence, IMLI is currently characterized by a striking lack of direction and it is often tolerated in the belief that it has few drawbacks. In fact, IMLI has commonly objectives that can be called auxiliary, i.e., directed at school achievement in other subjects. No attainment targets have been formulated with respect to proficiency in the immigrant minority languages in question. Target groups and enrolment The target groups of IMLI are not identified in terms of linguistic background, but in terms of country of birth, nationality, and the educational level of the parents. Immigrant minority policy is conceived of as a fight against inequality. A target pupil who is given educational priority from this perspective is a pupil • whose maternal grandmother was not born in Belgium and who is likewise not a Belgian national by birth, and • whose mother did not attend school beyond the end of the school year in which she turned 18. Most of the pupils who qualify for IMLI originate from Mediterranean countries, in particular from Turkey, Morocco, Italy, Spain, and Greece. In addition, Hebrew is taught to Jewish pupils at two Antwerp schools. Table 3.1 presents an overview of IMLI participation by primary schools and primary school pupils in the school year 1995/1996. Table 3.1 Participation in IMLI by primary schools and primary school children in 1995/1996 (Department of Education) Languages Number of schools Number of pupils Turkish 43 2.436 Arabic 18 957 Italian 15 520 Hebrew 2 228 Spanish 1 57 Greek 4 31 Aramaic 1 8 Total 84 4.237

< previous page

page_75

next page >

< previous page

page_76

next page > Page 76

At a rough estimate, this number is no more than 20% of the immigrant minority school population in Flanders. Neither in primary nor in secondary education is participation in IMLI open to indigenous Flemish pupils. Implementation and organization The language legislation in Flanders stipulates emphatically that Dutch is the language of instruction in education. Therefore, IMLI can be offered only as an optional program with experimental status that can be chosen in addition to compulsory education. A choice has to be made by the school board, teachers, and parents between a so-called 'supportive' or 'bicultural' model: • the supportive model stresses the auxiliary role of IMLI for learning other subjects; in this model, IMLI can be given up to a maximum of 20% of the available instruction time of 27 hours per week (i.e., 5.5 hours per week); • the bicultural model is directed at the acquisition and maintenance of active bilingualism in Dutch and the immigrant minority language in question; in this model, a maximum of 50% of the instruction time is available for IMLI. In fact, the bicultural model has been applied only in Brussels, which has an autonomous status with respect to Flanders and Wallonia. On the basis of the Brussels Foyer Project, bicultural educational experiments were carried out at some ten primary schools. The aim was for immigrant minority pupils to achieve trilingualism from kindergarten age onwards, the three languages being Dutch, French, and a particular immigrant minority language. An evaluation of the Foyer Project is presented in Byram and Leman (1990). At Flemish schools, the recruitment of IMLI teachers is subject to the following conditions: • the school must be funded or subsidized by the Flemish government; • the parents of the pupils in question must be informed in writing about the possibility to participate in IMLI; • at least two-thirds of the parents must give their written consent and at least 20 pupils must participate in IMLI at one school. The criteria apply to both primary and secondary education. These high thresholds partly account for the relatively limited participation in IMLI. In practice, usually less than half of the allowed 5.5 hours per week are devoted to IMLI at Flemish primary schools. In secondary education, there is a maximum of 2 hours per week. In secondary education, IMLI takes place both within and outside regular school hours. In the first case, it is either coupled to the contents of such subjects as geography or social education, or taught as an autonomous subject. It is almost exclusively for Moroccan pupils that IMLI is designed as an autonomous subject. In addition, IMLI is frequently offered outside regular school hours in secondary education, without government intervention, for Italian, Turkish, and Moroccan pupils.

< previous page

page_76

next page >

< previous page

page_77

next page > Page 77

Schools are free to take progress in immigrant minority language proficiency into consideration when evaluating school achievement. Only rarely do schools make use of this opportunity and give pupils an official mark for IMLI. In general, report marks for 'language' refer to progress in Dutch. IMLI in both primary and secondary education is generally funded by the embassies of the respective countries of origin. Only the above-mentioned Brussels Foyer Project has been partially paid for from the budget of the Flemish government, enabling it to recruit and professionalize a number of teachers of its own. In principle, schools can have the embassies reimburse them for the use of classrooms for IMLI. The embassies are also responsible for the selection and recruitment of IMLI teachers in the countries of origin. In actual fact, the teachers have very divergent - and often limited - professional qualifications. There are no standard procedures for the in-service training of IMLI teachers, and the quality of this training is often low. Teachers of Italian are given relatively good support facilities from Italy. Finally, the teaching materials for IMLI are often antiquated and frequently come from the countries of origin. So far, the Flemish government's policy with regard to IMLI has been directed at removing organizational obstacles for embassies. In 1994, the ministry made the Flemish Educational Council (VLOR) responsible for IMLI policy development. The construction of an outline curriculum and the setting of educational goals for Italian and Turkish should give new impetus to both the conceptualization and implementation of IMLI (see Snytsers 1994). This assignment runs parallel to action research being conducted by the universities of Ghent and Louvain, the aim of which is to describe a number of optimalization models at the school level. The outline curriculum referred to is as yet the subject of deliberation in the VLOR. 3.2.2 France2 French primary and secondary education consists of 2 years of nursery school (ages 4-5), 5 years of primary education (ages 610), 4 years of collège (ages 1114), and 4 years of lycée (ages 15-18). In primary education, IMLI is called ELCO (Enseignement des Langues et Cultures d'Origine); in secondary education, it is called ELV (Enseignement des Langues Vivantes). French IMLI policy is very restrictive. The historical reason for this can be traced to the French Revolution. The motto of the French Revolution, 'One state, one nation, one language,' not only devalued the status of regional minority languages and upgraded French as the language of national unity in the 19th and 20th century, but it also led to far-reaching centralism: educational policy in France is arranged 2 The section on IMLI in France is partly based on information of the Ministère de l'Education Nationale, the Inspection Generale de l'Education Nationale, the Institut National des Langues et Civilisations Orientales (INALCO) and the Centre National d'Enseignement à Distance (CNED). Special thanks are due to Dominique Caubet (Paris), Michel Neyreneuf (Paris) and Jan Jaap De Ruiter (Tilburg).

< previous page

page_77

next page >

< previous page

page_78

next page > Page 78

from Paris. In this centralized policy, the use of patois has historically been taboo. Cichon (1991), Ager (1993), and Caldwell (1994) give an overview of the status of regional minority languages in primary and secondary education. In primary education, German in Alsace, in particular, has maintained a relatively high status: in 1982, 75% of the pupils were enrolled in German as an optional subject for approximately 2 hours a week. Languages like Basque, Breton, Catalan, Corsican, and Occitan are only very rarely taught in primary schools. In secondary education, regional minority languages have occasionally been given an optional status as second or third modern languages, but enrolment has been low (Breton 3% in 1982-1987, Occitan 8% in 1985-1986). Both in primary and secondary education, minority languages are also languishing owing to a lack of continuity, teaching materials, and competent teachers. Regional minority languages are strongly marginalized by standard French in education; immigrant minority languages even more so. Both in primary and secondary education, immigrant minority pupils are strongly encouraged to learn French as a second language. Since the beginning of the eighties, IMLI has often been organized on the initiative and with the financial support of the consulates of the most important countries of origin, particularly Algeria, Tunisia, Morocco, Italy, Spain, Portugal, Turkey, and the former Yugoslavia. For studies on immigrants and their children in France, we refer to Tribalat (1995) and MARS (1996). Cassius (1994) gives an overview of French IMLI in general; Tilmatine (1997) and Masthoff (1998a) focus on IMLI for North African children. Arguments and objectives IMLI was initially established with a view to the remigration of migrant families to their countries of origin. This argument has been largely invalidated in the course of time. Other arguments have taken its place, which vary depending on who is being asked to argue the point. For the consulates of the Arabic countries of origin that fund this type of education (especially Morocco), IMLI is seen as an alternative to the religious courses that are propagated by Islamic organizations. Through IMLI, children acquire a minimum knowledge of the Arabic language and culture of their country of origin and can thus maintain ties with this country. A less overt argument is that these ties can then be subject to control by the consulates. Arabic parents often appear to have cultural and/or religious motives: children should learn Arabic to improve their knowledge of Arabic culture and/or the Koran. Finally, the Ministry of Education predominantly conceives of IMLI as conducive to integration and a successful school career, believing that immigrant minority pupils will be more successful if they are better acquainted with their community language and culture. Owing to the above-mentioned divergence in the arguments of various actors, reforms of cooperative agreements between France and the most important countries of origin are an uphill struggle. As a consequence, the objectives of IMLI in primary education remain very vague. Mixed groups of experts from France and the countries of origin are making an effort to establish language- and culture-

< previous page

page_78

next page >

< previous page

page_79

next page > Page 79

specific attainment targets. In secondary education, children are not instructed in their 'own' language, but they are incorporated into modern foreign language classes with attainment targets for language proficiency. Depending on the composition of their classes, teachers are expected to make allowance for differences in mothertongue backgrounds. Target groups and enrolment The participants in primary school IMLI are mainly children of parents from countries that fund this type of education through their consulates, even though IMLI is open to other children if the teacher involved and the school's headmaster give their consent. A survey of the languages that were available and chosen in the 1989/1990 school year is presented in Table 3.2. Table 3.2 Languages available and chosen in primary education 1989/1990 (Ministère de l'Education Nationale) Languages N N N pupils schools teachers Algerian Arabic 20.525 845 288 Moroccan Arabic 25.397 874 208 Tunisian Arabic 7.642 429 170 Portuguese 25.441 1.457 333 Turkish 16.398 721 193 Italian 12.572 337 74 Spanish 2.741 166 86 Yugoslavian 1.431 70 77 languages Total 112.1474.899 1.429 In the school year 1990/1991, approximately 30% of the Portuguese, Turkish, Italian, Spanish, and former Yugoslavian pupils enrolled in IMLI in primary education; less than 20% of the Arabic pupils did so. In secondary education, each pupil can take any foreign language as an optional subject. Table 3.3 presents figures for language enrolments in the school year 1991/1992. Some remarkable ethnocultural dividing lines are apparent in the choice of languages. Spanish is often chosen by French pupils, Portuguese much less, and Arabic hardly ever. Most pupils who enrol in Arabic as an optional subject are of Arabic origin. Standard Arabic is being taught in secondary education. Because this is a foreign language for most Arabophone pupils, their motivation is often rather low.

< previous page

page_79

next page >

< previous page

page_80

next page > Page 80

Table 3.3 Language enrolment in secondary education in 1991/1992 (Ministère de l'Education Nationale) Number of pupils Languages 5.150.702 English 1.444.906 Spanish 1.387.411 German 177.337 Italian 29.388 Russian 14.388 Portuguese 13.226 Standard Arabic 4.545 Hebrew 2.732 Chinese 1.403 Japanese 570 Dutch 297 Polish 4.121 Other languages 22.129 Regional languages 5.232 Correspondence courses Two alternative routes are open to anyone in France wishing to take exams in non-Western European languages that are not on the secondary education curriculum: enrolling in long-distance education at the Centre National d'Enseignement à Distance (CNED) or participating in the annual state exams without taking any courses to prepare for them. The state exams for nonWestern European languages are developed, organized, and assessed by INALCO (Institut National des Langues et Civilisations Orientales) in Paris. In 1995, INALCO administered written exams for a great number of non-Western European languages, including Albanian, Amharic, Arabic dialect, Bambara, Berber, Bulgarian, Cambodian, Croatian, Czech, Korean, Haussa, Hindi, Hungarian, Indonesian, Laotian, Lingala, Macedonian, Malagasy, Persian, Peuhl, Rumanian, Serbian, Slovakian, Slovenian, Swahili, and Tamil. In addition, Armenian, Finnish, Norwegian, Swedish, Turkish, and Vietnamese will also acquire the status of langues évaluables sous forme d'épreuves obligatoires. In 1995, a total of 11.878 people registered for state exams. The greatest number registered for an Arabic dialect (7.555 entries), followed by Berber (1.532), Turkish (600), Serbian (340), and Vietnamese (242). The written state exam for Arabic has two optional graphic variants (Arabic or Roman script) and two regional variants (Maghreb or Oriental Arabic). To avoid artificial languages, Maghreb languages have been optionally subdivided into Moroccan Arabic, Algerian Arabic, and Tunisian Arabic; Oriental Arabic has been subdivided into Egyptian, Syrian, Lebanese, and Palestinian Arabic (cf. Caubet 1997). Implementation and organization IMLI is offered in primary education if the consulate concerned wishes it and the Ministry of Education approves it. Often it is the parents who want IMLI and turn

< previous page

page_80

next page >

< previous page

page_81

next page > Page 81

to the consulate. There are no fixed guidelines for minimum enrolment: this is a matter of negotiations between the consulate, the Ministry, and the school management. The teachers are seconded to schools and paid by the consulates. In secondary education, a minimum of 5-8 participants is required for lessons to take place, depending on the type of secondary schooling. However, decisions to offer lessons are in fact not taken on the basis of numerical criteria alone, but are largely in the hands of school managements and boards, who have autonomous powers. School managements are often against it, either on budgetary grounds (it is easier and cheaper to offer just one or two foreign languages in school), or for more covert reasons (if Arabic is on the curriculum, this tends to attract immigrant minority rather than indigenous majority pupils to the school). The IMLI teachers are employed by the Ministry of Education and must, therefore, be French nationals. These are mostly teachers from Arab countries. With financial assistance from the consulates, 1-3 hours of IMLI per week can be taught in primary schools, either instead of other subjects, or outside regular school hours (in the afternoon or evening). Both variants are common. In the latter case, lessons usually take place in school classrooms anyway. In either case, IMLI is given during the entire primary school period or during the last four years. With the financial support of the Ministry of Education, 3-4 hours of IMLI per week are provided in secondary education, integrated into the general curriculum. For pupils from different educational sectors in the school, teaching sometimes takes place outside regular school hours (in the afternoon or evening). In primary education, attainment targets for evaluating immigrant minority language proficiency are lacking. An attempt is being made to integrate IMLI into the general curriculum and to evaluate it in terms of overall school achievement; sometimes a separate report mark is given. In secondary education, immigrant minority language proficiency is tested like any other subject. If candidates pass a state exam, the result may count as an additional credit on school reports. Primary school teachers are recruited through the consulates of the various countries of origin. The French government is not involved in their previous training or subsequent schooling, though IMLI teachers are eligible for participation in training programmes for other teachers. Some teachers participate, on an experimental basis, in in-service training programmes that are organized by the education inspectorate. Secondary school teachers are recruited through public selection procedures, and they may participate in the regular training programmes. The consulates are responsible for curriculum development in primary education, which is being done by mixed expert groups. The French government is responsible for curriculum development in secondary education, which in the case of IMLI chiefly concerns Arabic and Portuguese.

< previous page

page_81

next page >

< previous page

page_82

next page > Page 82

3.2.3 Germany3 Germany is a federal republic consisting of 16 autonomous states, each of which has its own responsibility and its own ministry of education. In the present federal republic, two minority groups have special language rights: the Danish-speaking inhabitants of Schleswig-Holstein and the Sorbs in Sachsen. The specific status of these language groups will not be discussed. It was not until the German unification that the themes of immigrant minority policy and IMLI were given a place on the social and political agenda in the former German Democratic Republic: Anderssein gab es nicht was Krüger-Portratz's (1991) conclusion with respect to the status of immigrants and minorities in the GDR. Arguments and objectives It is at the periodical Kultusministerkonferenz (KMK) that, among other things, the IMLI policies in the 16 federal states are attuned to one another. The most important recommendations of the Kultusministerkonferenz relating to IMLI can be summarized as follows (cf. Röhr-Sendlmeier 1986, Luchtenberg 1995, BAGIV 1985). In 1964, the decision was made to adopt a two-pronged educational policy with respect to immigrant minority children. The dual aim of such education was: (1) integration into the German educational system and, consequently, assistance in learning German, and (2) re-integration into the educational system of the country of origin, and therefore instruction in the mother tongue (muttersprachlicher Ergänzungsunterricht). The second aim was abandoned in 1971 as a result of actual demographic developments. The first aim has been in the foreground ever since. For learning German as a second language, so-called preparatory classes were recommended, in which children are placed for a maximum period of one or two years prior to receiving regular instruction. There are two IMLI models. In the first, the German government (i.e., the federal state) is responsible for the recruitment of teachers, the development of teaching materials, and the organization of IMLI. In the second model, the responsibility lies with the country of origin. In 1976, the Kultusministerkonferenz returned in its recommendations to the dual aims of 1964. It was now argued, with respect to IMLI, that the maintenance of the pupils' own identity was an important task of the school. The overriding argument for IMLI, however, remained the possibility of re-integration into the education system of the country of origin. What is finally worth mentioning here is the adaptation of 1979, which made it possible for pupils in secondary education to choose their mother tongue as a school subject instead of a first or second foreign language (KMK 1979). 3 The section on IMLI in Germany is partly based on interviews with the Sekretariat der Kultusministerkonferenz (Bonn), with the Landesinstitut für Schule und Weiterbildung (Soest, Nordrhein-Westfalen), and with the Regionale Arbeitsstelle zur Förderung ausländischer Kinder und Jugendlicher (Köln). Special thanks are due to Rob Aspenslagh (The Hague), Arnold Hock (Köln), Sigrid Luchtenberg (Dortmund), Eike Thürmann (Soest) and Hella Rottenberg (De Volkskrant).

< previous page

page_82

next page >

< previous page

page_83

next page > Page 83

In view of the responsibility of the federal states, the recommendations of the Kultusministerkonferenz do not constitute laws, although they are often at the root of legal measures. Thürmann (1994) and Luchtenberg (1995) point out that the rationale for IMLI in Germany is, or will be, subject to change. The focus will shift from a policy aimed at a return to the country of origin, to one in which the importance of multilingualism as a key qualification in a European context is stressed. Target groups and enrolment The starting point for determining the number of ausländische Schüler is their parents' nationality. In 1994, there were 1.122.010 immigrant minority pupils in German education (algemein bildende und berufliche Schulen), which was more than 9% of the total number of pupils. The two largest immigrant minority groups that could be distinguished on the basis of the parents' nationality were formed by 472.226 Turkish children and by 190.650 children whose origins are in the former Yugoslavia. In the first instance, IMLI in most federal states was limited to children belonging to families from the former recruitment countries for foreign workers (the so-called Entsendestaaten), as mentioned in the 1977 EU directive. These are the EU countries Italy, Greece, Spain, and Portugal as well as Turkey and the former Yugoslavia. In addition, some federal states offer IMLI to children from Morocco and Tunisia (KMK 1986). Mehlem (1998) gives an overview of Moroccan children in Germany. Masthoff (1998b) focuses on IMLI for Moroccan children in Germany. In a number of federal states, the legislation has recently been modified so that other immigrant minority children also qualify for IMLI. An example of this is Niedersachsen, where, since 1993, IMLI has been made possible for children from immigrant families from non-recruitment countries. In practice, this means that Farsi (Iran), Vietnamese, Kurdish, and Polish have become available in Niedersachsen as languages of instruction. Specific to Germany is the policy with respect to the so-called Aussiedler, i.e., people from Poland, the former Soviet Union, Roumania, and the former Yugoslavia who can prove that, via their (often remote) ancestors, they are of German descent. Whereas it is difficult for traditional Gastarbeiter and new immigrant groups such as asylum seekers or refugees to acquire German nationality, this is very easy for returned Aussiedler. Children from Aussiedler families are often given instruction in their non-German native language. In spite of both the limited school status and the extra time load, there is a keen interest in IMLI on the part of most immigrant minority groups. Roughly speaking, 60-80% of the pupils in primary and secondary education qualifying for IMLI actually participate in it (cf. Luchtenberg 1995). Table 3.4 presents an overview of participation in IMLI in the school year 1995/1996 in Nordrhein-Westfalen for the largest immigrant minority groups (cf. LSW 1996).

< previous page

page_83

next page >

< previous page

page_84

next page > Page 84

Table 3.4 Participation in IMLI in 1995/1996 in NordrheinWestfalen (LSW 1996) Country of origin N total N IMLI % Turkey 138.239 90.386 65 Former Yugoslavia 26.668 7.961 30 Italy 18.798 8.840 47 Bosnia and HG 13.114 unknown unknown Morocco 11.915 5.117 43 Greece 10.300 5.780 56 Croatia 4.418 2.800 63 Portugal 4.008 3.111 78 Spain 3.711 3.296 89 These data show that there are remarkable differences in participation in IMLI between different immigrant minority groups in Nordrhein-Westfalen. Participation in IMLI is highest (89%) for Spain as the country of origin and lowest (30%) for the former Yugoslavia (Croatia is mentioned separately). Implementation and organization In the former West German federal states, two models for IMLI obtain: • As far as content, organization, and funding are concerned, IMLI falls under the responsibility of the consulate of the country of origin. This is the model opted for by Baden-Württemberg, Berlin, Bremen, Hamburg, Saarland, and Schleswig-Holstein. • Responsibility for IMLI resides with the education department (Schulverwaltung) of the particular federal state. This model is operative in the federal states Bayern, Niedersachsen, Nordrhein-Westfalen, and Rheinland-Pfalz. Although the federal state of Hessen has also adopted this model, it occupies a special place in that IMLI has been compulsory since 1983 (see KochArzberger et al 1993). In the federal states of the former GDR (Brandenburg, MecklenburgVorpommern, Sachsen, Sachsen-Anhalt, and Thüringen), no IMLI measures have been taken comparable to those taken in the other federal states, an important reason being the relatively small number of immigrant minority pupils. Only in Sachsen have consulates been allowed, since 1992, to offer IMLI. IMLI based on the consulate model is becoming increasingly less popular, as it is increasingly difficult to integrate it into the regular programme. The fact that consulate-supervised IMLI is less easy to influence is also considered a serious drawback. In such specific conditions as those in which, for example, Serbian and Croatian pupils find themselves, control over the actual implementation of IMLI and the teachers' backgrounds is felt to be desirable. At the same time, the countries of origin are also less inclined to make the arrangements for IMLI themselves. Consulates of EU countries such as Portugal and Spain are, however, increasingly active in promoting knowledge about their countries' language and culture among EU citizens in Germany (cf. Böcker and Thürmann 1991).

< previous page

page_84

next page >

< previous page

page_85

next page > Page 85

Education is provided for a large number of immigrant minority languages in the various federal states. For example, the Kultusministerium of Nordrhein-Westfalen offers structural IMLI facilities for the following languages: Albanese, Arabic (for pupils from Morocco/Tunisia), Greek, Farsi (Iran), Croatian, Serbian, Italian, Polish, Portuguese, Russian, Spanish, and Turkish. De iure, it is usually the standard language of the country of origin which is chosen as the target language of IMLI; defacto, non-standard languages are also offered in a number of federal states. The need for IMLI is assessed by the school management on the basis of talks with the parents. Subsequently, the total need per school district is determined. The minimum number of pupils required per language varies from state to state. In the federal states with IMLI following the consulate model, there is usually no explicit mention of a minimum number of pupils. If such a minimum number is mentioned, it varies from 8 to 15 pupils per class (cf. Damanakis 1983). In the secondary schools of some federal states, it is possible to take an immigrant minority language as an optional subject instead of the first or second (obligatory) modern foreign language. Sometimes no lessons are given, the only option being to take an exam. Again there are substantial differences between the federal states: sometimes IMLI is offered only in a certain type of school (grammar school, vocational education); sometimes it is only available for pupils who have not attended German schools from their sixth year; sometimes it is only given in specific languages. In terms of availability and participation, the most progress has been made with Turkish as an optional subject in secondary education (see Kranemann and Hillebrand 1993, for an overview). From the perspective of increasing European cooperation, the importance of being able to learn a foreign language at an early age has been discussed in various federal states (cf. Luchtenberg 1995). In Nordrhein-Westfalen, for example, the concept of teaching 'neighbourhood' languages (Begegnungssprachen) is being discussed. Apart from English, these are not only French and Dutch in the areas bordering on France, Belgium, and the Netherlands, but also languages like Turkish, Italian, or Portuguese at schools with high concentrations of pupils speaking these languages as their mother tongues. In Rheinland-Pfalz and Baden-Württemberg, federal states bordering on France, there are also primary schools where pupils can receive bilingual instruction in German and French at an early age. Finally, some states have so-called European schools which specifically aim at a European dimension in their teaching (Artz and Cremer 1994:255-284). Berlin is the only state where this European conception is also adhered to in primary education; in the other federal states, it is only operative in secondary education. In these European schools, language instruction is an important component. However, immigrant minority languages are only marginally covered by this European dimension. The European schools in Berlin, which traditionally have paid attention to English, French, and Russian, have recently included Turkish in their curriculum.

< previous page

page_85

next page >

< previous page

page_86

next page > Page 86

In general, a maximum of five hours per week can be devoted to IMLI. In some federal states, an extra two hours of IMLI can be offered, but only if religion (particularly Islam or Greek Orthodoxy) is part of this instruction. In practice, the number of hours is probably lower, an estimated two and a half hours per week. In primary education, IMLI lessons are usually given during the school day; often some of the lessons (two hours) are taught parallel to religious instruction for the other pupils. Participation in IMLI is voluntary. Only in Hessen is IMLI participation obligatory, although it is quite easy for parents to obtain an exemption. The progress in language proficiency made by pupils who participate in IMLI in primary education is often expressed in a separate report mark. In Hessen, where participation is obligatory, the IMLI mark has the same status as the marks for other subjects and also carries weight in the end-of-year report. In the other federal states, the status of this report mark is limited and not comparable to the status of the marks for other school subjects; in some federal states, the status of the IMLI mark may be upgraded in the future. The financing and appointment of IMLI teachers varies again per federal state. In the consulate model, IMLI is largely funded by the countries of origin. Financial support by the federal state itself is usually restricted to the rental of classrooms. The teachers, too, are under the supervision of the country of origin. They come to Germany in rotation, being appointed for a maximum of five years. When IMLI is offered under the responsibility of the federal state, IMLI teachers are appointed by the Kultusministerium and often teach at different schools. Some of these teachers have passed further examinations and therefore gained a status similar to German teachers. However, IMLI funds are not earmarked as such and can also be employed for instruction in German as a second language to immigrant minority pupils. Recently, a new and so far unique teacher training program for Turkish (Studiengang Tiirkisch für das Lehramt) was established at Essen University. Obviously, consulate-funded IMLI is cheaper for the federal state. Gogolin (1993) gives a rough, but nonetheless informative estimate for Hamburg. In 1990, nearly 30.000 pupils (= 11%) had a non-German passport. Hamburg spends about DM 600.000 on IMLI annually. This amount would be at least ten times higher if IMLI did not fall under the responsibility of the consulates in Hamburg. As a result of the complex IMLI conditions, there are only a few federal states with clear directives for the IMLI curriculum. In federal states where the consulate model is operative, the content and aims of IMLI are generally determined by the countries of origin, where the teaching materials employed are usually developed. Since 1989, methodological-didactic recommendations for IMLI have been made available at the instigation of the Kultusministerkonferenz (KMK 1989, 1995). Some of the federal states themselves also play an active role in the development of educational plans and teaching materials. In this respect,

< previous page

page_86

next page >

< previous page

page_87

next page > Page 87

Hessen and Nordrhein-Westfalen deserve special mention (see, e.g., Kranemann and Hillebrand 1993). In the long run, IMLI policy in most federal states will be greatly changed. On the one hand, various documents and studies draw attention to the link between knowledge of languages and a strong position in a unified Europe. On the other hand, there is a serious mismatch between the directives in the various member states and actual IMLI practice. An important factor of change is the dynamics in the composition of immigrant minority groups, especially the sharp increase in residents from Eastern Europe. Another factor is the tendency towards increasing decentralization. For the time being, however, IMLI in Germany remains Ergänzungsunterricht and continues to have the unequal status of an optional and extra facility separate from the regular education programme. 3.2.4 Great Britain4 In this section, we will discuss the particular ways in which pupils from immigrant minority groups are being identified and the role of language in this process. Educational policy on Welsh in Wales and Gaelic in Scotland respectively will not be dealt with. The central government in London (Home Office) allocates funds to the local authorities of major districts and regions. As a rule, these local authorities are granted a great deal of liberty and flexibility in how they spend these funds. In education, however, their freedom is limited to a considerable extent by the constraints of the National Curriculum. All children between five and sixteen go through this compulsory curriculum, which consists of eleven subjects and is divided into four phases relating to the children's ages. The National Curriculum fully engages a pupil's school week. In phase 3, from age eleven, a modern foreign language is offered. All official languages of the European Union are considered modern foreign languages. The National Curriculum was introduced in 1988; it was evaluated in 1994 and will not be readjusted until the year 2000 (cf. DFEE 1995a, 1995b, 1995c). Educational facilities to promote racial equality in education (Bangs 1994) are funded mainly by special subsidies. A ministerial guideline known as Section 11 of the 1966 Local Government Act plays a key role here. This guideline allows an earmarked budget to be allocated to local authorities, who are to take special measures if a community includes persons belonging to immigrant minority groups whose language or customs differ from those of the rest of the community (Mackinnon, Statham and Hales 1995). 4 The section on IMLI in England is partly based on OFSTED (1994). There have also been interviews with Local Education Authorities for London and Berkshire, with the Department for Education and Employment (London) and with the Commission for Racial Equality. Special thanks are due to Glenn Adelman (Reading), Halima Field (London), Frances Migniuolo (London), and Ken Ponder (London).

< previous page

page_87

next page >

< previous page

page_88

next page > Page 88

Arguments and objectives Since the publication of the influential report entitled Education for All (DES 1985), the official government policy has been defined as follows: • to raise the performance level of all pupils and to remove the obstacles to higher achievement which are common to all; • to give immigrant minority pupils the same opportunity as all others to profit from what schools can offer them by meeting their particular educational needs, for example, by promoting good practice in the teaching of English; • to preserve and transmit the national values in a way which accepts Britain's ethnic diversity and promotes tolerance and racial harmony. Any attention paid to languages spoken by immigrant minority groups in addition to, or instead of, English almost exclusively serves an auxiliary educational objective: it serves to develop proficiency in English as a second language in order to enable immigrant minority children to participate in the National Curriculum successfully. Although the British government has stated that the ethnocultural diversity of British society should be reflected in the school curriculum, in practice, this boils down to intercultural projects, additional support provided in primary education by bilingual assistants or teachers in the classroom, and additional facilities in secondary education after the national languages of the European Union have been served first. In short, English education is predominantly education in English, and immigrant minority groups themselves are made responsible for organizing IMLI. Target groups and enrolment In language policy and education programmes for immigrant minority children, attention is mainly devoted to English as a second language and to intercultural education. Hence, the importance of specifying target groups for IMLI and data on IMLI enrolment has become marginal. Nevertheless, the definition and identification of immigrant minority groups has played an important role in minority policy and IMLI policy. Developments in the last thirty years show remarkable differences in the extent to which the careers of pupils from immigrant minority groups can be traced. The education system in Great Britain is largely decentralized. At the national level, the Department for Education and Employment (DFEE) has gathered information about the number of pupils from immigrant minority groups since 1966. At first, the birth-country of the pupils and their parents was used as a combined identification criterion. However, the usefulness and the validity of this statistical information met with such scepticism that the government decided to discontinue gathering these data in 1973 (DES 1986:5). As a result, only minimum information about the number and composition of immigrant minority pupil populations was available for a long time. At the local level, Local Educational Authorities (LEA) have a long-standing tradition and expertise with respect to the identification of pupils from immigrant

< previous page

page_88

next page >

< previous page

page_89

next page > Page 89

minority groups and the policy decisions on the organization of education to be taken on the basis of these identification data. LEAs have regularly held language surveys over the last decades (cf. Nicholas 1988). The Inner London Educational Authority (ILEA) has held such surveys at primary and secondary schools at two-year intervals since 1978. Since 1981, the focus of the language question has been on speaking 'a language other than, or in addition to, English at home.' Over time, a steady increase was found in the percentage of non-English home languages, ranging from 14% in 1981 to 23% in 1987. The number of different languages spoken also increased. In the last ILEA survey, held in 1989, 184 languages were mentioned, 16 of which were spoken by more than 1000 children. In addition to the LEA surveys, there have been complementary language surveys in London and in a wider variety of districts. One of the first relatively large-scale studies was done by Rosen and Burgess (1980). Their starting point was Fishman's (1965) classic question: 'Who speaks what languages to whom and when?' Teachers at 28 schools in Central London were asked to collect data about the use of languages other than standard English by their 11- and 12-year-old pupils. According to these teachers, 15%, in a population of 4.600 pupils, spoke a language other than English, and 15%, an overseas, mostly Caribbean, dialect of English. Rosen and Burgess' (1980) investigations and those of the ILEAs have yielded relevant data about the diversity of pupil populations. The educational implications of the language diversity observed are discussed by Edwards (1983), Houlton and Willey (1983), and Tansley, Navaz and Roussou (1985). However, these studies have also been criticized (see Nicholas 1988 and Alladina 1993): first, for their emphasis on determining proficiency in English; second, for failing to collect information about proficiency in the home language, especially the ability to read/write; third, for failing to map the need for instruction in the home language; and finally, for taking the knowledge of the teachers as a starting point in the data collection, with no guarantee that the information was obtained directly from the pupils. In the early eighties, a third language survey was commissioned by the Department of Education and Science (DES) and conducted in five districts, with the cooperation of London University. This Linguistic Minorities Project (LMP 1985) is one of the most comprehensive language inquiries ever conducted. Questionnaires and procedures were taken over by a large number of other LEAs. The criticism levelled at earlier language inquiries was taken into account by the LMP. For example, pupils were directly addressed as respondents by phrasing questions in the second person. The LMP consisted of three parts: • In the School Language Survey, the teacher put the following screening question to every pupil separately: 'Do you yourself ever speak any language at home apart from English?' Pupils who answered this question in the affirmative were subsequently asked three complementary questions: 'What is the name of that language? Can you read that language? Can you write that language?' The purpose of these questions was to provide insight into the

< previous page

page_89

next page >

< previous page

page_90

next page > Page 90

degree of language diversity at school and to give an indication of the skills in languages spoken at home other than English. • In the Secondary Pupils Survey, pupils had to fill out a questionnaire themselves. Detailed sociolinguistic information was gathered about where, why, and with whom, a certain language was spoken. • The Adult Language Use Survey, finally, was set up against the background of studies by Fishman (1965) and others. With the help of local bilingual experts, a bilingual questionnaire with 156 (!) questions was presented to 2.500 adults. A detailed picture was obtained of domains of language use, attitudes towards bilingualism, and processes of intergenerational language shift. In the Linguistic Minorities Project, the language use of more than 216.000 pupils was mapped. The number of pupils speaking a language other than English at home varied per district from 7% to 30%. The results of the project are discussed extensively in LMP (1985). In this report, attention is also paid to the implications of the project for the education of, and policies regarding, pupils of immigrant minority groups. In 1985, the influential Swann report Education for All (DES 1985) pointed out the difficulty of drawing up a budget for provisions in the absence of nationally recognized educational statistics. In the period 1990-1995, all schools in England were expected to carry out an annual Ethnic Monitoring Survey and to send off a survey of their total pupil population to the central Department of Education and Employment. Questions in the survey concerned ethnic origin, mother tongue, and religious affiliation. However, it was unclear how the ethnic origin of pupils should be interpreted. In the explanatory section for the schools, ethnic origin was considered to be a more or less objective feature of a pupil, although subjective self-categorization by the pupil was also taken into consideration. In the official registration of pupils' native languages in the school year 1994/1995 (DFEE 1995b:9), the following languages were distinguished: Bengali, Cantonese, Greek, Gujarati, Hindi, Italian, Punjabi, Portuguese, Spanish, Turkish, and Urdu. If more than 20 pupils in a school were categorized as Other, their mother tongues were also to be mentioned explicitly. The identification of pupils on the basis of the language criterion also raised some questions. Through the questionnaire, information was obtained about a pupil's mother tongue, whereas the schools had actually been instructed to collect data about the language or languages spoken in the home. Moreover, no information was obtained about the intensity with which a specific language was used, nor about the persons with whom it was used. The school obtained data on pupils' ethnicity by interviewing the parents on a voluntary basis. If the parents did not want to provide such information, the school was not entitled to make assumptions about ethnicity. The ethnicity data for those particular pupils would then be lacking and they were listed as unclassified. How information was gathered depended on the local situation. The most important options were: a mailed questionnaire for the parents, an interview

< previous page

page_90

next page >

< previous page

page_91

next page > Page 91

of the parents by the headmaster, and a group interview between the headmaster and several parents, in which each parent was asked to fill out a questionnaire. Because after a period of five years a great number of pupils were categorized as unclassified by the schools, the system of an annual Ethnic Monitoring Survey of education was thoroughly revised. In addition, the ethnicity data could no longer be linked to other educational indicators such as the level of achievement in the National Curriculum and the achievement of individual schools. As of the school year 1995/1996, the separate Ethnic Monitoring Survey described above was abandoned. Since then, information about the educational needs and school achievement of immigrant minority pupils has been obtained from a combination of three sources of information: the annual school census, the four-yearly evaluation by the education inspectorate (OFSTED), and periodical ad hoc surveys. The school census form, which is used to map the entire school population the 18th of January of each year being the reference date -, is very extensive. The information that is specifically obtained about immigrant minority pupils is, however, minimal; in the 1996 school census, such information concerned the following: • the total number of pupils that belong to an immigrant minority group; • the identification (not the numbers) of the four largest immigrant minority groups; • the total number of pupils for whom English is a second language. In contrast to the Ethnic Monitoring Survey, no information is requested about the pupils' mother tongue, and the pupils' knowledge of English as a second language has clearly gained prominence. The categorization into immigrant groups was carried over from the 1991 census. The schools are advised to collect the ethnicity data on the school census form chiefly from the parents. The Ministry of Education is aware that the data they obtain provide only a rough sketch of the multicultural composition and the language needs of a school. Therefore, additional information is collected by the education inspectorate at least once every four years (OFSTED 1994). This concerns information about: • the number of pupils in each major immigrant group; • the number of pupils for whom English is not their first language; • the four most important languages other than English spoken by pupils in the school; • the number of pupils eligible for so-called Section 11 support and the corresponding number of Section 11 teachers. The education inspectorate, therefore, does obtain some information about languages, but this is information about their use in school, not at home. Schools must adhere to the guidelines of the National Curriculum. In practice, this means that the curriculum allows little time for IMLI. In primary education, an immigrant minority language may be offered for auxiliary purposes only;

< previous page

page_91

next page >

< previous page

page_92

next page > Page 92

IMLI cannot be taught as an autonomous subject. In secondary education, the most important immigrant minority languages can be offered as subjects in the National Curriculum. Parents are to indicate their need for IMLI to be taught as a subject. Subsequently, at least two conditions have to be met: priority has to be given to national languages of the EU and sufficient funds have to be made available in order to be able, at the very least, to appoint an examination committee. IMLI is occasionally organized by immigrant groups themselves in, e.g., community centres. In this case, government funds are only available for accommodation. Implementation and organization To meet the specific needs of pupils from immigrant minority groups, bilingual assistants are appointed by some local authorities. These assistants are not considered educational staff, and they are generally funded by specially earmarked Section 11 budgets. There are also foreign-language assistants in secondary education, in addition to qualified teachers, to teach or assist pupils who have opted for a second foreign language on the curriculum, such as French or German. Proficiency in English is considered the key to access to, and success in, the National Curriculum. Within this narrow framework, local (education) authorities are responsible for the education of immigrant minority pupils. Many reports mention the government's intention to further decentralize educational policy, which means that the schools' budgets are directly transferred to the schools themselves, bypassing the local education authorities. However, many schools themselves, local education authorities, and immigrant minority groups do not welcome such far-reaching decentralization efforts. For a long time, the Section 11 target groups had been limited to immigrant minorities from Commonwealth countries like India and Pakistan. As of the school year 1995/1996, however, all immigrant minorities in Britain are in principle eligible for Section 11 support. These Section 11 funds can reach the local authorities in two ways: firstly, as an earmarked part of the total resourcing for local services (including funds for the fire brigade or the police). In addition to the number of pupils in a certain district, the so-called AEN factor (Additional Educational Needs) is also used to determine what educational facilities will be needed. This AEN factor consists of the number of: • children of single parents; • children of income support claimants; • residents under 16, born outside the Irish Republic, the USA, or the Commonwealth. Secondly, Section 11 funds can be obtained by the Local Education Authorities (LEAs) bidding for earmarked funding to support immigrant minority groups. LEAs can apply directly to the central government (Home Office) for such funds; however, there is no guarantee that all bids will be successful.

< previous page

page_92

next page >

< previous page

page_93

next page > Page 93

In addition, some LEAs have also successfully applied for grants from the Single Regeneration Budget (SRB) which is also provided by the central government. This SRB funding is intended to promote the general social and economic regeneration of communities, including the increasing of people's educational skills and training qualifications. Some of this funding is being used to support Section 11 type of activities in schools for immigrant minority pupils. Some SRB funding also supports IMLI activities by local immigrant minority organizations. Finally, LEAs can also apply directly to the central government for Grants for Education Support and Training (GEST). These grants must be spent in accordance with the terms of the National Curriculum. Since the school year 1995/1996, part of GEST has been aimed at developing teaching skills for dealing with linguistic diversity in the classroom and at promoting pupils' mastery of English, thus securing full access to the National Curriculum. The grants from the central government (whether via Section 11, SRB of GEST) all require the LEAs to provide matched funding in order to receive grants to cover between 40%60% of the total costs. 3.2.5 The Netherlands5. In the Netherlands, regional language varieties, in addition to the Dutch language, have traditionally been common phenomena. Frisian, however, is the only regional language variety with a legally protected status that extends to primary and secondary education. The introduction of Frisian into primary education was historically motivated by the argument of combatting deficiencies in school achievement (De Jong and Riemersma 1994). However, Frisian-speaking children are no longer less successful in education than Dutch-speaking children in Friesland. The motivation for instruction in Frisian has shifted from deficit arguments to cultural-political arguments in the course of time. In primary education, Frisian is currently a compulsory subject for all pupils (generally one hour a week); in secondary education, it is an optional subject. For a comparison of the status of instruction in Frisian with that of IMLI in the Netherlands, see Extra (1989); for a comparison of the status and use of various immigrant minority languages in the Netherlands, see Broeder and Extra (1995c, 1997a), Broeder et al (1993), and Extra and Verhoeven (1993a). Richters (1998) focuses on instruction in Arabic for Moroccan children in the Netherlands. The introduction of IMLI in the Netherlands has a remarkable history (see Lucassen and Köbben 1992 and Demirbas 1990 for a historical overview). In 1974, IMLI (Onderwijs in Eigen Taal en Cultuur or OETC in Dutch) was intro5The section on IMLI in the Netherlands is partly based on information of the Ministry of Education, Culture and Science, and the Inspectorate of Education. Special thanks are due to Marianne Nijdam (Ministry of Education, Culture and Science), and Seyfi Özgüzel (Inspectorate of Education).

< previous page

page_93

next page >

< previous page

page_94

next page > Page 94

duced for large groups of pupils in primary schools without any previous preparation, counselling, supervision, or set curriculum, and it even lacked any legal basis. Also in the Netherlands, IMLI was originally introduced into primary schools with a view to the remigration of migrant families. At the end of the seventies, however, the remigration argument faded into the background. IMLI developments should be understood against the backdrop of the developments in minority policy in the Netherlands after 1980. This policy was focused very strongly on socio-economic deficits; this perspective has also dominated educational policy for immigrant minority pupils. Schools with large numbers of pupils from socio-economically disadvantaged families were granted additional funds to combat deficiencies. Because immigrant minority children come from socioeconomically disadvantaged families more often than indigenous majority children, this deficit perspective was even reinforced. In its advisory report Ceders in de tuin ('Cedars in the garden,' CALO 1992), the Committee for Immigrant Minority Pupils in Education opted for a perspective in which immigrant minority languages are not primarily considered as sources of problems and deficiencies, but as sources of knowledge and enrichment. This latter starting-point requires a multicultural perspective rather than a deficit perspective. The perspective that is chosen has different consequences for one's choice of objectives, target groups, target languages, and evaluation of IMLI. Arguments and objectives After the remigration perspective had faded into the background, the objectives of IMLI in primary education were increasingly formulated in auxiliary terms. IMLI should contribute to bridging the gap between the home and school environment and to promoting school success in Dutch and other obligatory subjects. The number of those who have advocated intrinsic objectives, i.e., promoting first-language proficiency, is remarkably small. The National Institute for Curriculum Development (SLO 1996) has elaborated core objectives for eight curricular areas in primary education, including three languages, viz., Dutch, Frisian, and English. The core objectives of these languages concern oral and written skills and language awareness. Despite references to the development of the Netherlands into a multicultural society, core objectives for immigrant minority languages are lacking. Objectives in terms of mastery of Turkish and Arabic in primary and secondary education have been proposed by Diephuis et al (1992a) and Diephuis et al (1992b). It is a remarkable fact that such intrinsic objectives were accepted much sooner and on a much wider scale in secondary education (cf. the detailed achievement targets for Turkish and Arabic in lower vocational education and junior secondary education, as formulated by the National Examination Committee in Uitleg 1990). In Cedars in the garden, too, intrinsic rather than auxiliary objectives are primarily advocated, both in primary and in secondary education (CALO 1992).

< previous page

page_94

next page >

< previous page

page_95

next page > Page 95

Target groups and enrolment Which primary school immigrant minority pupils were eligible in the past for IMLI has been specified in school circulars (but not in legislation); these are • pupils of Moluccan origin; • pupils with at least one parent who originates from the former Mediterranean recruitment countries; • pupils with at least one parent who has been admitted as a refugee; • pupils from one of the EU member states. Funding for the target groups was subject to three kinds of restrictions: the target groups were exhaustively described in terms of a historically determined and limited list of countries of origin; provisions were meant to be temporary for first- and secondgeneration children; and disadvantaged socio-economic status was an important criterion for IMLI eligibility. Chinese children, for instance, have not been eligible in the past for IMLI because it has not been demonstrated that they suffer socio-economic disadvantages similar to those of other immigrant minority groups. On similar grounds, pupils from EU countries were likewise not eligible for IMLI funding. The Cedars in the garden report proposed dropping the generation criterion and the socioeconomic status criterion and replacing them with the home-language criterion: IMLI should be made available to those immigrant minority children who speak an immigrant minority language at home in addition to, or instead of, Dutch with at least one of the parents, and it should be supplied by municipalities and schools if there is a certain demand and if qualified teachers are available. In its response to Cedars in the garden, the Minister of Education endorsed the starting-points for IMLI laid down in this report, but proposed limiting the introduction of new target groups to Chinese, Antillean, and Surinamese pupils on organizational and financial grounds (cf. Ritzen 1993). Although this meant a partial acknowledgement of multicultural motives, the government opted once again for an extensive enumeration of target groups rather than for principled criteria. The choice of target languages has caused problems whenever the home language deviates considerably from the standard language of the country of origin. This is particularly so for Moroccan children, who often speak a Berber language at home. Many Moroccan parents, however, prefer instruction in Arabic for their children, on cultural and religious grounds. If the home language diverges widely from the standard language, parents and pupils are granted a right of choice in Cedars in the garden (under certain restrictive conditions), on the basis of the principle of cultural self-orientation. In primary education, such a choice of the home language instead of the standard language of the country of origin has already been made available to Moluccan children (Moluccan Malay) and to Syrian Orthodox children from Turkey (Aramaic). Table 3.5 presents an overview of IMLI enrolment figures of various groups of immigrant minority children. First of all, it shows that enrolment is subject to

< previous page

page_95

next page >

< previous page

page_96

next page > Page 96

both temporal and inter-group fluctuations. Enrolment of Turkish and Moroccan pupils remains relatively high, which points to a widely felt need. Such differentiation in time and in groups will be strongly determined by a combination of factors, such as the degree of institutional support from headmasters and school boards, the parents' and pupils' attitude towards IMLI, and the geographical distribution of immigrant minority groups (which will determine whether or not the numerical criterion of eight pupils per language group per school can be met). No comparative studies on the actual causes of this differentiated picture have been undertaken so far. Table 3.5 Enrolment of immigrant minority pupils in IMLI at elementary schools in 1990 and 1993 (Ministry of Education) 1990 1993 Country of origin N total N IMLI % N total N IMLI % Morocco 38.867 27.506 71 41.373 28.205 68 Turkey 38.294 31.328 82 42.619 33.002 77 Moluccan Islands 4.755 1.726 36 3.970 1.559 39 Former Yugoslavia 2.989 807 27 4.474 1.129 25 Spain 2.721 914 34 2.244 706 32 Italy 2.529 262 10 2.170 271 13 Cape Verde Islands 2.462 1.031 42 2.189 417 19 Portugal 1.506 508 34 1.890 1.095 58 Greece 815 318 39 855 261 31 Tunisia 671 69 10 969 298 31 Total 95.609 64.469 67 102.753 66.943 65 In secondary education, students with a non-Dutch cultural background may be taught the languages of their countries of origin as optional subjects. The long-term intention is to make IMLI available to all interested pupils, independent of their home language background. Table 3.6 shows the spread of pupils over the languages that were chosen in four consecutive school years. These figures make clear why Turkish and Arabic have a forefront position in the introduction of IMLI at Dutch secondary schools (see also Özgüzel 1994 and Aarssen, Broeder and Extra 1998). In 1995, 153 pupils took the nationwide written Turkish exam, and 53 pupils took the nation-wide written Arabic exam.

< previous page

page_96

next page >

< previous page

page_97

next page > Page 97

Table 3.6 Spread of pupils in secondary education over languages chosen in the period 1992-1996 (Ministry of Education) Languages chosen 1994/19951995/19961996/19971997/1998 Albanese 30 28 56 64 Arabic 3667 3601 3784 3644 Aramese 9 30 23 Chinese 26 English 28 30 Farsi 29 French 1 13 Hebrew 58 Italian 1 15 25 Moluccan/Malay 44 30 17 17 Portuguese 152 253 212 328 Russian 4 Serbian/Croatian/Bosnian313 249 207 222 Spanish 19 48 80 70 Turkish 3922 3746 3678 3530 Urdu 50 Vietnamese 19 24 15 Total 8.204 8.010 8.174 8.023 Implementation and organization The former status of IMLI was laid down in Article 10 of the Primary Education Act (WBO), which stipulated that a school with pupils with a non-Dutch cultural background might include IMLI in its curriculum if these pupils' parents desired it. Whether IMLI would actually be offered was determined by the outcomes of an interplay of forces of school managements, teachers, and parents. Article 10 of the WBO also stipulated that a maximum of 2.5 hours of IMLI per week might be provided within regular school hours and 2.5 hours outside school hours. In practice these stipulations have been interpreted differently. The Staff Establishment Decree WBO 1992 stipulated that the government would support IMLI in primary education for the above mentioned target groups by providing additional staff if there was a minimum of eight pupils per language group per school. Clusters of schools could also jointly draft an education plan, on the basis of which more pupils could be reached with the same budget. In secondary education, immigrant minority languages are available as optional subjects in conformity with Article 12 of the Secondary Education Act, either in addition to, or instead of, other optional subjects. To provide this kind of education, there is a provision for additional staff if a minimum of four pupils enrol for a certain language for two hours a week. Schools can jointly arrange this type of instruction for each other's pupils, which means that tasks can be divided on the basis of mutual agreements. The school offering a certain language receives compensation for all pupils enrolled for such instruction. Evaluative studies of IMLI in primary education indicate that there is a remarkable preference for examining the effects of IMLI on second-language

< previous page

page_97

next page >

< previous page

page_98

next page > Page 98

acquisition and/or school achievement rather than on first-language acquisition: first-language achievement is rarely considered as part of the evaluation. In the Cedars in the garden report, progress due to IMLI was emphatically considered as educational achievement that should be assessed. Empirical studies of the effect of IMLI on first-language proficiency are few and far between. Aarts, De Ruiter and Verhoeven (1993) and Aarts (1994) found that instruction in Turkish had positive effects on the Turkish proficiency of Turkish primary school pupils, whereas similar effects of instruction in Arabic were much less evident for Moroccan pupils (cf. Driessen 1990 for similar results). Meanwhile, at the request of the National Institute for Educational Measurement (CITO), tests have been developed by Aarts and De Ruiter (1995) that permit assessment of language proficiency in Arabic and Turkish at the end of the primary school period. Bilingual tests have also been developed for CITO by Verhoeven et al (1995) that allow the nature and degree of bilingualism to be determined at the beginning of the primary school period for Turkish, Moroccan, and Antillean pupils in Turkish and Dutch, Arabic and Dutch, and Papiamentu and Dutch, respectively. Finally, the requirements that IMLI teachers in primary schools had to meet in the past have been specified in the Qualifications Decree WBO. These requirements were strongly biased towards proficiency in Dutch (as a second language) and much less towards the didactic knowledge and skills needed for providing IMLI to young children. IMLI teachers are predominantly trained by means of refresher courses. Many IMLI teachers work as subject teachers in part-time positions along with group teachers in full-time positions; they often have to live up to many expectations at the same time, which, besides teaching, may include the development of (additional) teaching materials and intermediary contacts with parents and other schools. In the last few years, new curricula and teaching materials have been developed for specific languages, in particular for Moluccan Malay (SLO 1990), Arabic (SAC 1995), and Chinese (ABC 1995). For a well-documented overview of IMLI teaching materials, see Van Dijk and Lodeweges (1992). The importance of local responsibilities is being increasingly emphasized in the government's national education policy. Hence, a new balance in the responsibilities of government, municipalities, and schools is being pursued. This pursuit is the central theme of the memorandum Local Education Policy by State Secretary of Education Netelenbos (1995a). This memorandum proposed decentralizing the responsibilities and means for school accommodation, counselling, prioritization, Dutch as a second language, and IMLI to the local councils of municipalities. This view was further elaborated in the memorandum Education in Immigrant Minority Languages (Onderwijs in Allochtone Levende Talen, curtly OALT) by State Secretary Netelenbos (1995b), which focused on primary education. Important elements of this memorandum were the recognition of the widely felt need for IMLI; the appreciation of the personal and social importance of knowledge of immigrant minority languages; the acceptance of

< previous page

page_98

next page >

< previous page

page_99

next page > Page 99

government responsibility for the provision and the quality of IMLI; and the acknowledgement of the importance of the home language criterion rather than the socio-economic status criterion or the generation criterion for eligibility for IMLI. All these elements have been adopted from the above-mentioned report Cedars in the garden. A new element was the emphasis on local language policy. In the State Secretary's conception, the government remains responsible for: • a national policy framework (which is not further elaborated); • earmarked funding of IMLI (to prevent improper use of allocated funds); • national quality control of IMLI (by the inspectorate of education). The municipal councils are responsible for: • information about IMLI and IMLI needs assessment; • selective allocation of IMLI funds to schools on the basis of proposals for IMLI submitted by schools; • inter-school cooperation on IMLI for minor language groups; • the involvement of immigrant minority groups as actors rather than as target groups only in establishing local IMLI policy. Finally, the schools remain responsible for: • the implementation and quality of IMLI; • the recruitment and employment of IMLI teachers. In December 1997, the State Secretary's policy proposals on IMLI were approved by the Dutch parliament. It remains to be seen how the new legislation, which has come into force in August 1998, will be implemented in practice (see also Broeder and Extra 1996a, 1996b). Kabdan, Robijns, and Van Eck (1997) present different organizational models for local IMLI policy. As yet, the contents of the national policy framework remain vague. The distant and noncommittal tone may lead to inequalities in IMLI provisions and to a decrease rather than an increase in the legitimacy of IMLI. Furthermore, means for suiting practical actions to principled words are limited: although earmarking of IMLI funds is actually pursued, so are fixed budgets and liberty of spending. An auxiliary and an intrinsic function of IMLI are also distinguished, which, in at least two policy proposals, may lead to a negative representation of the intrinsic function: the auxiliary function of IMLI alone can be realized within regular school hours (the intrinsic function is to be realized outside school hours), and only this auxiliary function is embedded in quality and training requirements for teachers. In the public debate on the implementation of the new IMLI legislation, fears have been expressed that the extra-curricular status of IMLI will lead to its marginalization and that the responsibility for its content will be given away to non-educational special interest groups. Recent developments in Sweden (see the next section) also show that a decentralized government policy need not necessarily benefit the position of IMLI.

< previous page

page_99

next page >

< previous page

page_100

next page > Page 100

3.2.6 Sweden6 Information about IMLI in Sweden is relevant for a number of reasons. Sweden has always made facilities for IMLI available in the context of a relatively strongly decentralized government policy (see Boyd 1993 and Hyltenstam and Arnberg 1988, for an overview). An historic landmark in government policy on IMLI is the Home Language Reform (Hemspråksreform) of 1976, which was approved by the Swedish parliament. Unlike other continental European countries, no sizeable labour migration from Mediterranean recruitment countries affected Sweden in the sixties and seventies. As the concept of remigrating migrant-labour families was lacking, the debate about IMLI and its implementation has been more straightforward in Sweden. In addition, as a consequence of political and economic changes, the responsibility for, and the implementation of, IMLI was increasingly decentralized in the eighties. A comparative history of Swedish and continental European immigration policies is outlined by Hammar (1981). Lainio (1997) focuses on the special status of Finnish-speaking pupils in Swedish schools. Arguments and objectives Immigration policy in Sweden has three central objectives, which were accepted by the Swedish parliament in 1975 and reconfirmed in 1986 (SO 1991): • equality (Jämlikhetsmål): the standard of living for immigrants should be equal to the rest of the Swedish population; • freedom of choice (Valfrihetsmål): immigrants should have the opportunity to determine for themselves to what extent they wish to maintain or develop their original cultural and linguistic identity, and to what extent they wish to be absorbed into Swedish society; • cooperation (Samverkansmål): mutual understanding, tolerance, and solidarity between immigrants and the Swedish population should be aimed at in decision-making on immigrant issues. These three objectives of Swedish immigrant policy led to the Home Language Reform (Hemspråksreform) in 1976, which legalized the immigrant minority pupils' obligation to participate in instruction in Swedish as a second language and their right to instruction in a language other than Swedish that is spoken at home. In organizing and improving education for immigrant minority pupils, language is the main focus (Hyltenstam and Tuomela 1996, LPO 1994). The most important objective is the acquisition of oral and written bilingual skills and the development of a strongly bicultural identity (LPO 1994). In stressing cultural 6 For the section on IMLI in Sweden special thanks are due to Sally Boyd (Göteborg), Kenneth Hyltenstam (Stockholm), and Peter Hedberg (Orebro).

< previous page

page_100

next page >

< previous page

page_101

next page > Page 101

arguments and bilingual objectives, Sweden distinguishes itself from most other continental European countries. Target groups and enrolment The target groups of IMLI are not defined in terms of (low) socio-economic status, but in terms of home language. It concerns the following groups: • pupils with one or two parents or guardians speaking a home language other than Swedish and who use this home language with their parents or guardians every day; • pupils who speak Swedish as their home language but who come from foreign schools; • immigrant pupils who speak predominantly Swedish at home and ask for instruction in the language of the country of origin; • pupils belonging to national minority groups (chiefly Lapps and Finns); • pupils of Romanes origin; • adopted children with a mother tongue other than Swedish. The Swedish education system consists of nine years of primary education (Grundskola) for 7- to 16-year-olds, and three (formerly two to four) years of secondary education (Gymnasieskola) According to home language statistics, in October 1990, 12% (103.390) of the pupils in Grundskola and 7% (20.290) of the pupils in Gymnasieskola spoke a language other than Swedish at home. All in all, about 130 different languages were spoken, Finnish, Spanish, and Arabic being the front-runners. 19% of the pupils with a home language other than Swedish spoke no or hardly any Swedish; 65% (66.900) of the pupils in Grundskola with a home language other than Swedish were enrolled in IMLI, as were 50% of these pupils in Gymnasieskola. For roughly 40 out of the 130 home languages, no IMLI of any kind was available, mainly due to a lack of qualified teachers or because the minimum number of pupils for a particular language within a municipality was not met. In Sweden, boards of primary schools are legally required to carry out the following activities every year: • make a survey of home languages other than Swedish spoken by any of the pupils in the school; • provide information for parents of non-Swedish-speaking pupils - both in Swedish and in the most important local immigrant minority languages - about the importance and the possibility of IMLI; • assess these parents' needs of IMLI for their children; • provide IMLI for a minimum of 5 pupils per language group per municipality (which requires schools to cooperate if demand falls below this minimum). This last numerical criterion is more generous than that of any other European country in this study. Table 3.7 presents a nationwide overview (1994) of eligibility for, and relative enrolment in, IMLI and Swedish as a second language

< previous page

page_101

next page >

< previous page

page_102

next page > Page 102

in primary education for language groups with more than 500 pupils entitled to these facilities. Bosnian, Croatian, and Serbian are reported as one group, given the fact that many people from former Yugoslavia no longer refer to their language as SerboCroatian. Table 3.7 Eligibility for and enrolment in IMLI and Swedish as a second language in primary education in 1994 (Swedish Bureau of Statistics 1995) Languages Eligibility Enrolment % Enrolment in % in IMLI L2 Swedish Albanese 2.838 1.255 442.123 75 Arabic 11.688 7.625 658.341 71 Assyrian 2.609 1.389 531.905 73 Bosnian/Croatian/Serbian 11.488 5.387 478.423 73 Cantonese 889 606 68559 63 Chinese 876 454 52512 58 Danish 1.023 151 15231 23 German 1.430 582 41211 15 English 4.537 2.454 541.340 30 Finnish 18.877 9.906 527.667 41 French 1.012 539 53374 37 Greek 2.145 1.371 64923 43 Hungarian 1.497 598 40491 33 Italian 809 398 49152 19 Kurdish 2.745 1.592 582.054 75 Macedonian 665 271 41283 43 Norwegian 630 41 7 107 17 Persian 8.330 5.994 724.988 60 Polish 5.141 2.728 531.775 35 Portuguese 813 395 49370 46 Roumanian 816 266 33388 48 Russian 857 400 47510 60 Somalian 2.059 1.348 651.803 88 Sorani 617 511 83513 83 Spanish 9.631 6.075 635.058 53 Syrian 918 607 66651 71 Thai 836 290 35522 62 Tigrinese 1.516 1.024 681.054 70 Turkish 4.846 2.967 613.566 74 Vietnamese 918 494 54746 81 The relative enrolment in IMLI and Swedish as a second language shows a great many mutual differences. Table 3.8 contains the total data on primary and secondary education in 1994.

< previous page

page_102

next page >

< previous page

page_103

next page > Page 103

Table 3.8 Eligibility for and enrolment in IMLI in primary and secondary education, autumn 1994 (Swedish Bureau of Statistics 1995) Pupils Grundskola Gymnasieskola Total number of pupils 916.661 309.952 Possible eligibility for IMLI 111.720 19.583 Actual enrolment in IMLI 61.306 (=55%) 7.160 (=37%) In view of the fact that IMLI is not compulsory and often an extra-curricular activity, these enrolment figures are relatively high. Nevertheless, in the mid-eighties, they had been considerably higher (70% and 50% for primary and secondary education, respectively). Implementation and organization A child can enrol in IMLI for one language only, even if several languages are spoken at home in addition to or instead of Swedish. IMLI is not open to children who do not use the immigrant minority language in everyday life. There is still a clearcut distinction between IMLI for children who speak a language other than Swedish at home and traditional language education. This is clearly exemplified by the fact that IMLI teachers of German are employed at other institutes and receive other kinds of training than teachers of German as a foreign language (cf. Reich 1994). In contrast to most continental European countries, the parents' mother tongue is the language that should be offered in education de iure. However, the national standard language is often offered de facto. In many - though not all - immigrant minority groups, the parents have a preference for instruction in the non-standard home language. The parents' wishes carry a great deal of weight. Swedish schools offer separate IMLI for Serbians, Croatians, and Bosnians; children from Kurdish families receive instruction in Kurdish if the parents so desire. In determining the criteria for eligibility for IMLI, the rules are bent more for some groups than for others. Indigenous minority groups, like the Saami and the Tornedal Finns, or children from Romanes families, can also enrol in IMLI if their home language is not Swedish. Moreover, IMLI will often be made available to these groups if there are less than five potential participants in a municipality. Every year, information is obtained about all school pupils for the nation-wide home language statistics (hemspråkstatistik). The schools send their filled-out forms to the Department of Educational Statistics of the Central Bureau of Statistics. The home language statistics provide an important basis for Swedish educational policies; they have direct implications for the organization and funding of education. The data are presented annually by the CBS under the title Hemspråk och hemspråksundervisning ('Home language and home language instruction'). The headmaster or teacher collects the following data on each individual pupil, with the aid of the parents if necessary: (1) the language used at home in addition to, or instead of, Swedish (only one language per pupil may be

< previous page

page_103

next page >

< previous page

page_104

next page > Page 104

reported); (2) the need for, and enrolment in, IMLI; (3) the need for, and enrolment in, instruction in Swedish as a second language. The headmaster of a school must report the required number of teaching hours in Swedish as a second language to the municipal education department. In addition, the following information must be specified for each home language: (1) the required number of teaching hours per week; (2) the required number of home language classes; and (3) the estimated number of pupils in home language classes (a home language class being a class in which all pupils have a common home language other than Swedish). Variables 2 and 3 were collected for the last time in 1991. When the National Agency for Education (Skolverket) was created, these variables were considered too expensive to collect or superfluous. The education inspectorate monitors the schools' and councils' compliance with legal obligations to provide IMLI where necessary. This is not to say, however, that the implementation of these rules and regulations in Sweden always proceeds smoothly (cf. Boyd 1993). Among the difficulties are the increasing decentralization of educational policy, compensatorily determined attitudes towards IMLI, the lack of qualified or available IMLI teachers, and the virtual absence of attainment targets and evaluations of IMLI. In the majority of cases, IMLI instruction is scheduled in the afternoon outside regular school hours and generally takes two hours a week. Pupils may enrol in IMLI for a maximum of 7 out of the 9 compulsory years in their school career. Another kind of IMLI is organized in bilingual schools. In this case, target pupils are in group 1-6 of Grundskola. The Finnish-speaking pupils may continue in groups 7-9. In bilingual schools, the home language largely functions as the instructional language. A minimum requirement is that half - or more in the upper grades - of the instruction is provided in Swedish. At the beginning of the eighties, around 15% of the pupils who spoke a language other than Swedish at home attended bilingual schools, but this is currently only 5%. This decrease was predominantly caused by declining numbers of Finnish-speaking pupils in schools and the reduction in budgets earmarked for bilingual education. In 1994, legal arrangements were made to allow for alternative ways of organizing IMLI. First of all, IMLI can be provided in Grundskola for 1 hour a week, starting from group 3 (to a maximum of 470 hours for the entire Grundskola period). In group 7 of Grundskola (age 13), pupils can opt for an immigrant minority language instead of a second foreign language (usually German or French). They will then receive three hours of IMLI a week (to a maximum of 320 hours for the remaining Grundskola period). A condition for such classes to take place is once again that there should be a minimum of five pupils in a municipality who wish to enrol and that there are IMLI provisions in secondary education. Little is known about enrolment in this new type of IMLI, but seeing that it has only recently been introduced, enrolment is still likely to be limited. In Sweden, pupils in primary schools do not get a mark for 'language' but for Swedish; apart from this mark, their progress in immigrant minority language

< previous page

page_104

next page >

< previous page

page_105

next page > Page 105

proficiency is assessed. Children get such marks for their school achievements from group 8 onwards. Immigrant minority pupils who are still enrolled in IMLI at this stage are usually given a separate report mark for immigrant minority language proficiency. The status of this last assessment in determining a pupil's overall school achievement varies widely in different educational forms. In Gymnasieskola, pupils can opt for an immigrant minority language as their second (or third or fourth) foreign language. The home language statistics that are collected annually are the most important data for local decision-making processes on educational policy in Sweden; they have immediate consequences for the organization and funding of IMLI. The central government supplies the local municipalities with funds. The role of the central government is limited to carrying out evaluative studies of school achievement, in particular cost-benefit analyses (for example, see Löfgren 1991). Initially, IMLI funds were earmarked per pupil through the home language statistics. Lately, however, local authorities have gained more and more freedom to spend. At present, IMLI funds are no longer earmarked within the decentralized educational budgets. Abandoning the system of earmarked funds has actually led to cuts in IMLI that exceed those in other school subjects. IMLI was therefore more badly hit than the educational provisions for Swedish as a second language. 3.3 Conclusions and Discussion In all the countries involved in this study, there has been an increase in the number of immigrant minority pupils who speak a language at home other than or in addition to the dominant school language in primary and secondary education. The schools have responded to this home-school language mismatch by paying more attention to the learning and teaching of the national standard language as a second language. A great deal of energy and money is being spent on developing curricula, attainment targets, teaching materials, and tests for second-language education. Instruction in the immigrant minority languages stands out in stark contrast to this, as it is much more susceptible to an ideological debate about its legitimacy. While there is consensus about the necessity of investing in second-language education for immigrant minority pupils, there is a lack of such support for IMLI. Immigrant minority languages are commonly considered sources of problems and deficiencies, and they are rarely seen as sources of knowledge and enrichment. Policy makers, headmasters, and teachers of 'regular' subjects often have reservations or are negative towards IMLI. On the other hand, parents of immigrant minority pupils, IMLI teachers, and immigrant minority organizations often make a case for having immigrant minority languages in the school curriculum. These differences in topdown and bottom-up attitudes emerge in all the countries focused upon in this study.

< previous page

page_105

next page >

< previous page

page_106

next page > Page 106

From a historical point of view, most of the countries in this study show a similar chronological development in their argumentation for IMLI. IMLI was generally introduced into primary education with a view to family remigration. In the seventies, this argumentation was virtually abandoned. Demographic developments showed no substantial sign of families remigrating to their former recruitment countries; instead, a process of generation forming and minorization came about. This development resulted in a shift and IMLI became aimed at combatting disadvantages. IMLI had to bridge the gap between the home and school environment and to encourage school achievement in 'regular' subjects. Because such an approach tended to underappreciate ethnocultural dimensions, a number of countries began to emphasize the intrinsic importance of knowledge of immigrant minority languages from a cultural, legal, and economic perspective: • in cultural respects, IMLI can contribute to maintaining and advancing a pluralist society; • in legal respects, IMLI can meet the internationally recognized right to language development and language maintenance, in correspondence with the fact that many immigrant minority groups consider their own language of key value to their cultural identity; • in economic respects, finally, immigrant minority languages and cultures can be an important pool of knowledge in a society that is increasingly internationally oriented. The historical development of arguments for IMLI in terms of remigration, combatting deficiencies, and cultural policy is particularly evident in the Netherlands, Germany, and Belgium. In France and Great Britain, cultural policy is tied in with the respective national languages French and English to such an extent that IMLI is only tolerated in its margins. In contrast to each of these five countries, cultural-political motives have always taken pride of place in Sweden. It should, however, be stressed that cultural-political arguments for IMLI have not led to an educational policy in which the status of immigrant minority languages has been substantially advanced in any of the countries involved in this study. The target groups of IMLI are considered disadvantaged groups in virtually all the countries in this study; only Sweden has an explicit home language criterion rather than a socio-economic status or generation criterion for admission to IMLI. Actual enrolment in IMLI varies widely not only between countries (cf. the enrolment percentages in the Netherlands versus Flanders), but also between groups (cf. the enrolment percentages of Moroccan and Turkish pupils versus those of Southern European pupils). Variation in enrolment is determined by a combination of factors, such as the attitudes of immigrant minority parents and pupils, and indigenous majority headmasters and teachers, and the geographical distribution of immigrant minority groups (which will decide whether or not numerical criteria can be met). As yet, comparative studies on the actual causes of this differentiated picture are not available.

< previous page

page_106

next page >

< previous page

page_107

next page > Page 107

There are remarkable differences in status between IMLI in primary and secondary education in the countries of this study. A comparison of target groups, arguments, objectives, evaluation, enrolment restrictions, curricular status, funding, and teaching materials shows that IMLI in secondary education has gained a higher status than IMLI in primary education. In primary education, IMLI is generally not part of the 'regular' or 'national' curriculum, and, consequently, it tends to become a negotiable entity in a complex and often opaque interplay of forces by several actors, in contrast with other curricular subjects. These differences are summarized in Table 3.9. Table 3.9 Status differences between IMLI in primary and secondary education Immigrant Primary education Secondary education Minority Language Instruction • de iure: mostly all pupils Target • de iure: mostly groups • defacto: mostly subset of immigrant minority immigrant minority pupils pupils from specific source countries • defacto: mostly subset of these pupils Arguments mostly in terms of a mostly cultural policy: struggle against deficits: • promoting cultural pluralism • bridging the • promoting knowledge of home/school gap languages • promoting school success in other ('regular') subjects rarely cultural policy: • promoting cultural pluralism • promoting knowledge of languages in a multicultural and globalizing society Goals rarely specified goals to commonly specification of oral be reached with IMLI and written skills to be reached with IMLI Evaluation rarely judgement/report examination and report figure for figure for IMLI: IMLI: national standard language 'language' in school is explicitly referred to and report = national separately evaluated in school standard language report Minimal relatively high number relatively low number of pupils: enrolment of pupils: specified per specified per class, school or class, school or municipality municipality Time-table not perceived as 'regular' regular optional subject in regular status education: instead of free time-table space other subjects or at extra-curricular hours Funding by national, regional or local • by national, regional authorities or local authorities • by consulates/embassies of source countries Educationalrarely originating from commonly originating from appliances country of settlement, country of settlement often from abroad/source country

< previous page

page_107

next page >

< previous page

page_108

next page > Page 108

The higher status of IMLI in secondary education is largely due to the fact that instruction in one or more languages other than the national standard language is a traditional and regular component of the (optional) school curriculum. Within secondary education, however, IMLI must compete with languages that, in their turn, have a higher status or a longer tradition. The hierarchy of languages in secondary education is schematically represented in Table 3.10 in descending order of status (1-5).

Table 3.10 Hierarchy of languages in secondary education, in descending order of status (15) With regard to category 5, it should be noted that some countries provide instruction and/or exams in non-standard language varieties. In France, for instance, pupils can take part in examinations for several varieties of Arabic and Berber (see Tilmatine 1977); Dutch primary education provides instruction in Moluccan Malay (as an alternative to Indonesian), and Sweden offers Kurdish (as an alternative to Turkish). Another remarkable fact is that in some countries (particularly France, Flanders/Belgium, and some German federal states), IMLI in primary education is funded by the consulates or embassies of the countries of origin concerned. In these cases, the national government does not interfere in the organization of IMLI, or in the requirements for, and the selection and employment of teachers. A paradoxical consequence of this phenomenon is that the earmarking of IMLI budgets by the above-mentioned consulates or embassies is often safeguarded. National, regional, or local governments often fail to earmark budgets, so that funds meant for IMLI are not infrequently appropriated for other educational purposes.

< previous page

page_108

next page >

< previous page

page_109

next page > Page 109

IMLI may be part of a largely centralized or decentralized educational policy. In the Netherlands, national responsibilities and means are gradually being transferred to the local level. In France, government policy is strongly centrally controlled. Germany has devolved governmental responsibilities chiefly to the federal states with all their mutual differences. Sweden grants farreaching autonomy to municipal councils in dealing with tasks and means. In England, there is a mixed system of shared national and local responsibilities (cf. the ministerial guidelines for special target groups versus the guidelines of the local educational authorities). In general, comparative cross-national references to experiences with IMLI in the various EU member states are rare (e.g., Reich 1991, 1994; Reid and Reich 1992; Fase 1994; Tilmatine 1997; Broeder and Extra 1997b, 1997c) or they focus on particular language groups (e.g., Tilmatine 1977; Obdeijn and De Ruiter 1998). With a view to the demographic development of these states into multicultural societies and the similarities in IMLI issues, more comparative research and cross-national policy initiatives would be desirable. Language policy is still developed within the national perspectives of the different EU member states. Proposals for a common EU language policy are labouriously achieved and noncommittal in character (see Coulmas 1991). The most important declarations, recommendations, or directives on language policy, each of which concepts carry a different charge in the EU jargon, concern the recognition of the status of (in the order mentioned): • national EU languages; • indigenous or regional minority languages; • immigrant or 'non-territorial' minority languages The Treaty of Rome (1958) confers equal status on all national languages of the EU member states (with the exception of Irish and Luxembourgian) as working languages. On numerous occasions, the EU ministers of education have declared that the EU citizens' knowledge of languages should be promoted (see Baetens Beardsmore 1993). Each EU member state should promote pupils' proficiency in at least two 'foreign' languages, and at least one of these languages should be the national standard language of one of the EU states. Promoting knowledge of regional and immigrant minority languages has been left out of consideration in these ministerial statements. The European Parliament accepted various resolutions in 1981, 1987 and 1994, in which the protection and promotion of regional minority languages was recommended. The first resolution led to the foundation of the European Bureau for Lesser Used Languages in 1982. Meanwhile, the Bureau has member state committees in 13 EU countries and it has recently acquired the status of Non-Governmental Organization (NGO) at the levels of the European Council and the United Nations. Another result of the European Parliament resolutions is the foundation of the European MERCATOR Network, aimed at promoting research on the status and use of regional minority languages. In March 1998, the European Charter of Regional

< previous page

page_109

next page >

< previous page

page_110

next page > Page 110

Minority Languages came into operation. This Charter was framed by the Council of Europe in 1992 and it has meanwhile been ratified by seven member states. The Charter is aimed at the protection and the promotion of regional minority languages, and it functions as an international instrument for the comparison of legal measures and other facilities of the EU member states in this policy domain. As yet, no such initiatives have been taken in the policy domain of immigrant minority languages. It is remarkable that the teaching of indigenous or regional minority languages is generally advocated for reasons of cultural diversity as a matter of course, whereas this is rarely a major argument in favour of teaching immigrant minority languages. In various EU countries, the 1977 guideline of the Council of European Communities on education for immigrant minority children (Directive 77/486, dated 25 July, 1977; see Appendix 3) has promoted the legitimization of IMLI and occasionally also its legislation (see Reid and Reich 1992, Fase 1987). The application of this guideline in establishing target groups for IMLI has nowhere been as restrictive as in Germany (see BAGIV 1985, KMK 1986). In Sweden, this guideline has never had any effect, as Sweden has only recently joined the EU. Meanwhile, the guideline needs to be reformulated and extended to pupils from non-EU countries, and it needs to be given greater binding force in the EU member states. The increasing internationalization of pupil populations in European schools, finally, requires a language policy for all pupils in which the traditional dichotomy between foreign language instruction for indigenous majority pupils and home language instruction for immigrant minority pupils is put aside.

< previous page

page_110

next page >

< previous page

page_111

next page > Page 111

References Aarssen, J, Broeder, P. and G. Extra (1998). Allochtone talen in het voortgezet onderwijs. Bouwstenen voor lokaal taalbeleid. Den Haag: VNG Uitgeverij. Aarssen, J., De Ruijter, J. and L. Verhoeven (1992). Toetsing Turks en Arabisch aan het einde van het basisonderwijs. Tilburg: Tilburg University Press, Studies in Meertaligheid 1. Aarssen, J. (1996). Relating events in two languages. Acquisition of cohesive devices by Turkish-Dutch bilingual children at school age. Tilburg: Tilburg University Press, Studies in Multilingualism 2. Aarts, R. (1994). Functionele geletterdheid van Turkse kinderen in Turkije en in Nederland. De Lier: Academisch Boeken Centrum. Aarts, R. and J. De Ruiter (1995). Toets Turkse Taal, Toets Arabische Taal. Arnhem: CITO. Aarts, R., De Ruiter, J. and L. Verhoeven (1993). Tweetaligheid en schoolsucces. Relevantie en opbrengst van etnische groepstalen in het basisonderwijs. Tilburg: Tilburg University Press, Studies in Meertaligheid 4. ABC (1995). Chinese taalmethode voor het basisonderwijs. Amsterdam: ABC Ager, D. (1993). Identity, community and language policies in contemporary France. In: D. Ager, Muskens, G. and S. Wright (eds), Language education for intercultural communication. Clevedon: Multilingual Matters. Alladina, S. (1993). South Asian languages in Britain. In: G. Extra and L. Verhoeven (eds), Immigrant languages in Europe. Clevedon: Multilingual Matters, 55-65. Artz, V. and W. Cremer (1994). Lernen für Europa. Neue Horizonte der Pädagogik. Bonn: Bundeszentrale für politische Bildung. Australian Institute of Multicultural Affairs (1982). Evaluation of post-arrival programs and services. Canberra: Australian Government Publication Service. Backus, A. (1996). Two in one. Bilingual speech of Turkish immigrants in the Netherlands. Tilburg: Tilburg University Press, Studies in Multilingualism 1. Baetens Beardsmore, H. (1993). European models of bilingual education. Clevedon: Multilingual Matters. BAGIV (1985). Muttersprachlicher Unterricht in der Bundesrepublik Deutschland. Sprach- und bildungspolitische Argumente für eine zweisprachige Erziehung von Kindern sprachlicher Minderheiten. Tagungsakten der Evangelischen Akademie Rheinland-Westfalen in Mülheim/Ruhr. Hamburg: Rissen. Bangs, J. (1994). Funding for race equality in education: looking for the future. Multicultural Teaching 13, 7-13. Böcker, E. and E. Thürmann (1991). Muttersprachlicher Unterricht für eine mehrsprachige und multikulturelle Schule. Der Deutschunterricht 43, 70-81. Boeschoten, H. (1990). Acquisition of Turkish by immigrant children. A multiple case study of Turkish children in the Netherlands aged 4 to 6. Wiesbaden: Harrasowitz. Bos, P. (1997). Development of bilingualism. A study of school-age Moroccan children in the Netherlands. Tilburg: Tilburg University Press, Studies in Multilingualism 8. Boumans, L. (1998). The syntax of codeswitching. Analysing Morroccan Arabic/Dutch conversation. Tilburg: Tilburg University Press. Studies in Multilingualism 12.

< previous page

page_111

next page >

< previous page

page_112

next page > Page 112

Boyd, S. (1993). Immigrant minority languages and education in Sweden. In: G. Extra and L. Verhoeven (eds), Immigrant languages in Europe. Clevedon: Multilingual Matters, 273-283. Broeder, P. and G. Extra (1994). Minority languages at Dutch elementary schools. Empirical data and educational policy. Journal of Intercultural Studies 15 (2), 53-71. Broeder, P. and G. Extra (1995a). Taal en etniciteit. De positie van allochtone talen in gezin en op school. In: H. Van den Eerenbeemt and J. Goedegebuure (eds), Cultuur en Identiteit. Tilburg: Tilburg University Press, 137-157. Broeder, P. and G. Extra (1995b). Ethnic identity and community languages in the Netherlands. In: H. Haarmann (ed), Sociolinguistica. International Yearbook of European Sociolinguistics 9: European identity and language diversity. Tübingen: Max Niemeyer, 96-112. Broeder, P. and G. Extra (1995c). Minderheidsgroepen en minderheidstalen. Den Haag: VNG Uitgeverij. Broeder, P. and G. Extra (1996a). Minderheidsgroepen en minderheidstalen in vergelijkend perspectief. Gramma/TTT 5 (2), 7797. Broeder, P. and G. Extra (1996b). Van OET naar OALT. De paarse zoektocht naar meertalig basisonderwijs. De Europese Gemeente 31, 11-17. Broeder, P. and G. Extra (1997a). Minority groups and minority languages in the Netherlands: empirical facts and educational policy. In: T. Bongaerts and K. De Bot (eds), Perspectives on foreign-language policy. Amsterdam: Benjamins, 157-179. Broeder, P. and G. Extra (1997b). Taal. In: H. Vermeulen (ed), Immigrantenbeleid voor de multiculturele samenleving. Integratie-, taal- en religiebeleid voor immigranten in vijf West-Europese landen. Amsterdam: Het Spinhuis, 54-94. Broeder, P. and G. Extra (1997c). Language. In: H. Vermeulen (ed), Immigrant policy for a multicultural society. A comparative study of integration, language and religious policy in five Western European countries. Brussels: Migration Policy Group, 57100. Broeder, P. and G. Extra (1998a). Migration and multilingualism in Western Europe. The Netherlands as case study. In: G. Extra and J. Maartens (eds), 141-160. Broeder, P. and G. Extra (1998b). Minderheidsgroepen en minderheidstalen in vergelijkend perspectief. In: G. Meijnen (ed), Opvoeding, onderwijs en sociale integratie. Groningen: Wolters-Noordhoff, 44-66. Broeder, P., Extra, G., Habraken, M., Van Hout, R. and H. Keurentjes (1993). Taalgebruik als indicator van etniciteit. Een studie naar identificatie van allochtone leerlingen. Tilburg: Tilburg University Press, Studies in Meertaligheid 3. Broeder, P., Extra, G. and J. Maartens (1998). Durban Language Survey. In: G. Extra and J. Maartens (eds), 121-137. Broeder, P., Extra, G. and R. Van Hout (1995). Language-related criteria as determinants of ethnicity. Goals and results of a feasibility study in the Netherlands. In: W. Fase, Jaspaert, K. and S. Kroon (eds), The state of minority languages. International perspectives on survival and decline. Amsterdam/Lisse: Swets and Zeitlinger, 113-134. Byram, M. and J. Leman (1990). Bicultural and trilingual education. Clevedon: Multilingual Matters. Caldwell, J. (1994). Provision for minority languages in France. Journal of Multilingual and Multicultural Development 15 293310.

< previous page

page_112

next page >

< previous page

page_113

next page > Page 113

CALO (Commissie Allochtone Leerlingen in het Onderwijs) (1992). Ceders in de tuin. Naar een nieuwe opzet van het onderwijsbeleid voor allochtone leerlingen. Delen 1 en 2. Den Haag: SDU. Cassius, M. (1994). Guide practique ELCO. Enseignement de Langue et Culture d'Origine. Nantes: CRDP. Caubet, D. (1997). L'épreuve facultative d'arabe dialectal du Bac français: passage à l'écrit. Bilan comparatif des sessions 1996 et 1995. In: M. Tilmatine (ed), 163-172. CBS (1995). Allochtonen in Nederland 1995. Voorburg/Heerlen: CBS. CBS (1997). Allochtonen in Nederland 1997. Voorburg/Heerlen: CBS. Charry, E., Koefoed, G. and P. Muysken (eds) (1983). De talen van Suriname. Achtergronden en ontwikkelingen. Muiderberg: Coutinho. Cichon, P. (1991). Sprachenpolitik und sprachliche Minderheiten in den Europaïschen Gemeinschaften. Das Beispiel Frankreich. In: S. Kroon et al (eds), Kultur- und Sprachenvielfalt in Europa. Münster: Waxmann, 71-87. Clyne, M. (1982). Multilingual Australia. Melbourne: River Seine Publ. Clyne, M. (1991). Community languages. The Australian experience. Cambridge: Cambridge University Press. Coulmas, F. (1991). A language policy for the European Community. Prospects and quandaries. Berlin: Mouton De Gruyter. Cummins, J. and M. Danesi (1990). Heritage languages. The development and denial of Canada's linguistic resources. Toronto: Garamond Press. Dales, I. (1990). Definitieve regeringsreactie op WRR-advies Allochtonenbeleid. Den Haag: Ministerie van Binnenlandse Zaken. Dales, I. (1992). Registratie en rapportage minderhedenbeleid. Den Haag: Staatsuitgeverij. Damanakis, M. (1983). Muttersprachlicher Unterricht für ausländische Schüler. Deutsch Lernen 4, 15-47. Damen, M. (1998). La communauté Flamande de Belgique. In: H. Obdeijn and J. De Ruiter (eds), 129-170). De Abreu, R. (1994). Surinaamse en Antilliaanse/Arubaanse leerlingen in het basisonderwijs in Tilburg-Noord. Een inventariserend onderzoek. Master thesis, Tilburg: Tilburg University. De Jong, S. and A. Riemersma (1994). Taalpeiling yn Fryslan. Onderzoek naar de beheersing van het Fries en het Nederlands aan het einde van de basisschool. Leeuwarden: Fryske Akademy. De Ruiter, J. (ed) (1991). Talen in Nederland. Een beschrijving van de taalsituatie van negen etnische groepen. Groningen: Wolters-Noordhoff. De Vries, J. (1985). Some methodological aspects of self-report questions on language and ethnicity. Journal of Multilingual and Multicultural Development 6, 347-368. De Vries, J. (1989). On coming to our census. A layman's guide to demolinguistics. In: D. Gorter et al (eds), Fourth International Conference on Minority Languages. Clevedon: Multilingual Matters, 57-76. De Vries, J. and F. Vallee (1980). Language use in Canada. Ottawa: Ministry of Supply and Services. Demirbas, N. (1990). OETC-beleid. Theorie en praktijk. Utrecht: NCB. DES (1985). Education for all. Report of the committee of inquiry into the education of children from ethnic minority groups. London: HMSO (The Swann Report).

< previous page

page_113

next page >

< previous page

page_114

next page > Page 114

DES (1986). Working Group on the collection of educational statistics on an ethnic basis. London: Department of Education and Science. DFEE (1995a). The English education system. An overview of structure and policy. London: Department for Education and Employment. DFEE (1995b). Ethnic monitoring of school pupils. A consultation paper. London: Department for Education and Employment. DFEE (1995c). The school curriculum. London: Department for Education and Employment. Dialektopoulos, T. (1998). Grieks onderwijs in eigen taal en cultuur in Nederland. Utrecht: LIZE. Diephuis, R., Galjee, W., Hajer, M., Meestringa, T. and H. Mulder (1992a). Kerndoelen Turks in het basisonderwijs en de eerstefase van het voortgezet onderwijs. Enschede: SLO. Diephuis, R., Hajer, M., Meestringa, T. and H. Mulder (1992b). Kerndoelen Arabisch in het basisonderwijs en de eerstefase van het voortgezet onderwijs. Enschede: SLO. Directive 77/486 (1977). Directive of the Council of the European Communities on the schooling of children of migrant workers. Brussels: CEC. Driessen, G. (1990). De onderwijspositie van allochtone leerlingen. De rol van sociaal-economische en etnisch-culturele factoren met speciale aandacht voor het OETC. Nijmegen: ITS. Edwards, V (1983). Languages in multicultural classrooms. London: Batsford. El Aissati, A. (1997). Language loss among native speakers of Moroccan Arabic in the Netherlands. Tilburg: Tilburg University Press, Studies in Multilingualism 6. EuroStat (1997). Migration statistics 1996. Statistical document 3A. Luxembourg: EuroStat. Extra, G. (1989). Ethnic minority languages versus Frisian in Dutch primary schools. A comparative perspective. Journal of Multilingual and Multicultural Development 10, 59-72. Extra, G. (1996). De multiculturele samenleving in ontwikkeling. Feiten, beeldvorming en beleid. Tilburg: Tilburg University Press. Extra, G. and J. Maartens (eds) (1998). Multilingualism in a multicultural context. Case studies on South Africa and Western Europe. Tilburg: Tilburg University Press, Studies in Multilingualism 10. Extra, G. and T. Vallen (1997). Migration and multilingualism in Western Europe. A case study of the Netherlands. In: W. Grabe (ed), Multilingualism. Cambridge: Cambridge University Press, 151-169 (Annual Review of Applied Linguistics 17). Extra, G. and L. Verhoeven (eds) (1993a). Community languages in the Netherlands. Amsterdam: Swets and Zeitlinger. Extra, G. and L. Verhoeven (eds) (1993b). Immigrant languages in Europe. Clevedon: Multilingual Matters. Extra, G. and L. Verhoeven (eds) (1998). Bilingualism and Migration. Berlin: Mouton De Gruyter. Fase, W. (1987). Voorbij de grenzen van onderwijs in eigen taal en cultuur. Meertaligheid op school in zes landen verkend. Rotterdam: Erasmus University. Fase, W. (1994). Ethnic divisions in Western European education. Münster/New York: Waxmann. Fernandes Mendes, H. (1991). Concept-nota registratie en rapportage minderhedenbeleid. Den Haag: Staatsuitgeverij.

< previous page

page_114

next page >

< previous page

page_115

next page > Page 115

Fishman, J. (1965). Who speaks what language to whom and when? Linguistics 2, 67-88. Fishman, J. (1989). Language and ethnicity in minority sociolinguistic perspective. Clevedon: Multilingual Matters. Fishman, J., et al (1985). The rise and fall of the ethnic revival. Berlin: Mouton De Gruyter. Giacolone Ramat, A. and P. Ramat (1998). The Indo-European languages. London: Routledge. Gogolin, I. (1993). Multilingualism in the German school. In: D. Ager, Muskens, G. and S. Wright (eds), Language education for intercultural communication. Clevedon: Multilingual Matters, 91-107 Grimes, B. (ed) (1996). Ethnologue. Languages of the world. Dallas: Summer Institute of Linguistics (also at http://www.sil.org/etnologue/). Grobler, E., Prinsloo, K. and P. Van der Merwe (eds) (1990). Language atlas of South Africa. Language and literacy patterns. Pretoria: HSRC. Habraken, M. (1992). Taalgebruik en taalattitudes van Surinaamse leerlingen als etniciteitscriteria. Master thesis, Tilburg: Tilburg University. Hammar, T. (1981). Swedish and European immigration policy - a comparative study. Stockholm: Swedish Commission on Immigration Research. Haskey, J. (1991). The ethnic minority population residents in private households. Estimates by country and metropolitan districts of England and Wales. Population Trends 61. Hendriks, H. (1986). De tweetaligheid van Chinese kinderen in Nederland. Master thesis, Leiden: Dept. of Sinology. Home Language Reform (1976). Norrköping: Statens Invandrarverk. Houlton, D. and R. Willey (1983). Supporting children's bilingualism. York: Longman for the Schools Council. Hyltenstam, K. and L. Arnberg (1988). Bilingualism and education of immigrant children and adults in Sweden. In: C. Bratt Paulston (ed), International Handbook of bilingualism and bilingual education. New York: Greenwood Press. Hyltenstam, K. and V. Tuomela (1996). Hemspråksundervisningen. In: K. Hyltenstam (ed), Tvåspråkighet med förhinder? Ivandrar- och minoritetsundervisning i Sverige. Lund: Studentlitteratur. ILEA (1990 fourth ed). Catalogue of languages spoken by ILEA school pupils. London: ILEA Research and Statistics Branch. Inspectie van het Onderwijs (1987). Onderwijs in eigen taal en cultuur in het voortgezet onderwijs. Inspectierapport 15. Den Haag: Staatsuitgeverij. Kabdan, R., Robijns, M. and E. Van Eck (1997). Van OET naar OALT. Organisatiemodellen voor lokaal beleid. Amsterdam: SCO. Kipp, S., Clyne, M. and A. Pauwels (1995). Immigration and Australia's language resources. Canberra: Australian Government Publication Service. Kipperman, B. (1994). Taalgebruik bij Cambodjaanse leerlingen in het basisonderwijs. Master thesis, Tilburg: Tilburg University. Klarberg, M. (1982). Foreign language literacy in Australia 1911, 1921 and 1933. Journal of Intercultural Studies 3, 91-5. Klatter-Folmer, J. and S. Kroon (eds) (1997). Dutch overseas. Studies in maintenance and loss of Dutch as an immigrant language. Tilburg: Tilburg University Press, Studies in Multilingualism 9.

< previous page

page_115

next page >

< previous page

page_116

next page > Page 116

KMK (1979). Unterricht für Kinder ausländischer Arbeitnehmer. Bonn: Sekretariat der Kultusministerkonferenz. KMK (1986). Zweiter Bericht der Bundesrepublik Deutschland über Richtlinie 77/486/EWG des Rates der Europäischen Gemeinschaften vom 25. Juli 1977 über die schulische Betreuung der Kinder von Wanderarbeitnehmern im Schuljahr 1984/85. Bonn: Sekretariat der Kultusministerkonferenz. KMK (1989). Grundlagen und methodisch-didaktische Empfehlungen für den muttersprachlichen Unterricht ausländischer Schüler. Bonn: Sekretariat der Kultusminister-konferenz. KMK (1995). Ausländische Schüler und Schulabsolventen 1985 bis 1994 Bonn: Sekretariat der Kultusministerkonferenz. Koch-Arzberger, C., Böhme, K., Hohmann, E. and K. Schacht (1993). Einwanderungsland Hessen? Daten, Fakten, Analysen. Bonn: Westdeutscher Verlag. Kook, H. (1994). Leren lezen en schrijven in een tweetalige context. Antilliaanse en Arubaanse kinderen in Nederland. PhD thesis, Amsterdam: University of Amsterdam. Kranemann, N. and U. Hillebrand (1993). Türkisch - ein neues Fach an Schulen in NordRhein-Westfalen. Schulverwaltung NRW 10, 232-235. Krige, D., Cairns, S., Makalima, B. and D. Scott (1994). The education atlas of South Africa. Durban: Education Foundation. Kriiger-Portratz, M. (1991). Ausländer und Minderheiten in der DDR. Anderssein gab es nicht. Münster: Waxmann Verlag. Lainio, J. (1997). Swedish minority language treatment and language policy. Positive public rhetoric versus grassroot struggle. In: U. Ammon, Mattheier, K. and P. Nelde (eds), Sociolinguistica. International Yearbook of European Sociolinguistics (11). Tübingen: Max Niemeyer Verlag, 29-42. Lieberson, S. (1981). Language diversity and language contact. Stanford: Stanford University Press. Linssen, B. (1994). Taalgebruik en taalvaardigheid van Turkse basisschoolleerlingen. Een studie naar thuistaalgebruik en taalvaardigheid van Turkse basisschoolleerlingen in Eindhoven. Master thesis, Tilburg: Tilburg University. LMP (Linguistic Minorities Project) (1985). The other languages of England. London: Routledge and Kegan Paul. Löfgren, H. (1991). Immigrant students in Sweden. A comparative study between immigrant groups and a sample of Swedish students. Malmö: University of Lund, Educational and Psychological Interactions 109. LPO (1994). Läroplanen för det obligatoriska skolväsendet och de frivilliga skolformerna. Stockholm: Department of Education. LSW (1996). Statistische Veröffentlichungen. Soest: Landesinstitut für Schule und Weiterbildung. Lucassen, L. and A. Köbben (1992). Het partiële gelijk. Controverses over het onderwijs in de eigen taal en cultuur en de rol daarbij van beleid en wetenschap (1951-1991). Amsterdam/Lisse: Swets and Zeitlinger. Luchtenberg, S. (1995). Perspectives on bilingual education in Germany. Journal for Intercultural Studies 16, 73-84. Luüs, C. and R. Oberholzer (1994). Provincial characteristics of South Africa. Johannesburg: Southern Book Publishers. Maartens J. (1998). Multilingualism and language policy in South Africa. In: G. Extra and J. Maartens (eds), 15-36.

< previous page

page_116

next page >

< previous page

page_117

next page > Page 117

Mackinnon, D., Statham, J. and M. Hales (1995). Education in the UK. Facts and figures. London: Hodder and Stoughton. MARS (1996). Le Monde Arabe dans la Recherche Scientifique 6. Paris: CNRS. Martens, E., Roijen, J. and J. Veenman (1994). Minderheden in Nederland. Statistisch Vademecum 1993/1994. Den Haag: SDU/CBS. Masthoff, P. (1998a). La France. In: H. Obdeijn and J. De Ruiter (eds), 17-70. Masthoff, P. (1998b). L'Allemagne. In: H. Obdeijn and J. De Ruiter (eds), 213-254. McDermott, L. (1998). The case of English in South Africa. In: G. Extra and J. Maartens (eds), 105-120. Mehlem, U. (1998). Zweisprachigkeit marokkanischer Kinder in Deutschland. Frankfurt/M: Peter Lang. Meloen, J. and J. Veenman (1990). Het is maar de vraag... Onderzoek naar de responseffecten bij minderhedensurveys. Lelystad: Vermande. Mesthrie, R. (ed) (1995). Language and social history. Studies in South African sociolinguistics. Cape Town: David Philip. Michielsens, M. (1992). Italia mia. Televisie kijken in ballingschap. Amersfoort/Leuven: Acco. Ministerie van Binnenlandse Zaken (1983). Minderhedennota. Den Haag: Staatsuitgeverij. Narain, G. (1995). Taaltalent in ontwikkeling. Een studie naar het Papiamentu en het Nederlands in de kleuterperiode op Curaçao en in Nederland. Tilburg: Tilburg University Press. Netelenbos, T. (1995a). Lokaal onderwijsbeleid. Zoetermeer: Ministerie van Onderwijs, Cultuur en Wetenschappen. Netelenbos, T. (1995b). Uitwerkingsnotitie onderwijs in allochtone levende talen. Zoetermeer: Ministerie van Onderwijs, Cultuur en Wetenschappen. Nicholas, J. (1988). British language diversity surveys (1977-1987). A critical examination. Language and Education 2, 15-33. Nicholas, J. (1992). Language diversity surveys as agents of awareness raising and change in education. PhD thesis, London: University of London. Nieuwenhuysen, J. (1993). Community profiles 1991 census: Netherlands born. Canberra: Bureau of Immigration and Population Research. Nortier, J. (1989). Dutch and Moroccan Arabic in contact. Code switching among Moroccans in the Netherlands. Dordrecht: Foris. Obdeijn, H. and J. De Ruiter (eds) (1998). Le Maroc au coeur de l'Europe. L'Enseignement de la langue arabe aux élèves marocains dans les pays d'accueil. Leiden: Centre pour l'histoire des migrations. OFSTED (1994). Educational support for minority ethnic communities. A survey of educational provision under Section 11 of the 1966 Local Government Act. London: Office for Standards in Education. Omzendbrief 13 BA/CD (1996). Gewoon basisonderwijs. Onderwijsbeleid voor migranten. Maatregelen voor het schooljaar 1996-1997. Brussel: Departement Onderwijs. Özgüzel, S. (1994). De vitaliteit van het Turks in Nederland. PhD thesis, Tilburg: Tilburg University. Ozolins, U. (1993). The politics of language in Australia. Cambridge: Cambridge University Press.

< previous page

page_117

next page >

< previous page

page_118

next page > Page 118

Reich, H. (1991). Developments in ethnic minority language teaching within the European Community. In: K. Jaspaert and S. Kroon (eds), Ethnic minority languages and education. Amsterdam/Lisse: Swets and Zeitlinger, 161-174. Reich, H. (1994). Unterricht der Herkunftssprachen von Migranten in anderen europäischen Einwanderungsländern. In: A. Dick (ed), Muttersprachlicher Unterricht. Ein Baustein für die Erziehung zur Mehrsprachigkeit. Wiesbaden: Hessisches Kultusministerium, 31-46. Reid, E. and H. Reich (1992). Breaking the boundaries. Migrant workers' children in the EC. Clevedon: Multilingual Matters. Research Group Kurds (1996). Koerden in Nederland: moedertaal en identiteit. Amsterdam: Dept. of Languages and Cultures of the Near East. Richters, J. (1998). Les Pays-Bas. In: H. Obdeijn and J. De Ruiter (eds), 171-212. Rinsampessy, E. (1992). Saudara Bersaudara. Molukse identiteit in processen van culturele verandering. Assen: Van Gorcum. Ritzen, J. (1993). Naar meer schoolsucces. Kabinetsreactie bij het advies 'Ceders in de tuin'. Zoetermeer: Ministerie van Onderwijs en Wetenschappen. Roelandt, Th., Roijen, J. and J. Veenman (1991). Minderheden in Nederland. Statistisch Vademecum 1991. Den Haag: Staatsuitgeverij. Röhr-Sendlmeier, U. (1986). Die Bildungspolitik zum Unterricht für ausländische Kinder in der Bundesrepublik Deutschland. Eine kritische Betrachtung der vergangenen 30 Jahren. Deutsch Lernen 1, 51-67. Romaine, S. (ed) (1991). Language in Australia. Cambridge: Cambridge University Press. Rosen, H. and T. Burgess (1980). Languages and dialects of London school children. London: Ward Lock Education. SAC (1995). Verder met Arabisch. Utrecht: SAC. Schaufeli, A. (1991). Turkish in an immigrant setting. A comparative study of the first language of monolingual and bilingual Turkish children. PhD thesis, Amsterdam: University of Amsterdam. Sillitoe, K. (1987). Developing questions on ethnicity and related topics for the census. London: OPCS. Silitoe, K. and P. White (1992). Ethnic group and the British census: the search for a question. Journal of the Royal Statistical Society 155, 141-163. SLO (1990). Leerplan onderwijs in eigen taal en cultuur in het basisonderwijs. Moluks Maleis. Enschede: SLO/Legio. SLO (1996). Kerndoelen basisonderwijs. Veldadvisering. Enschede: SLO. Smolicz, J. (1980). Language as a core value of culture. Journal of Applied Linguistics 11, 1-13. Smolicz, J. (1992). Minority languages as core values of ethnic cultures. A study of maintenance and erosion of Polish, Welsh, and Chinese languages in Australia. In: W. Fase, Jaspaert K. and S. Kroon (eds), Maintenance and loss of minority languages. Amsterdam: Benjamins, 277-305. Snytsers, F. (1994). Raamcurriculum OETC. Brussel: Vlaamse Onderwijsraad. SÖ (1991). Invandrare och flyktingar i det svenska utbildningssamhället (Immigrants and refugees in the Swedish educational society). Stockholm: National Board of Education.

< previous page

page_118

next page >

< previous page

page_119

next page > Page 119

Strating, H. (1994). Taalgebruik van Molukse leerlingen als identificatiecriterium. Een studie naar taalgebruik en taalvaardigheid van Molukse leerlingen in het basisonderwijs. Master thesis, Tilburg: Tilburg University. Swedish Bureau of Statistics (1995). Hemspråk och hemspråksundervisning. Örebro: SCB. Tahitu, E. (1989). Melaju Sini. Het Maleis van Molukse jongeren in Nederland. PhD thesis, Leiden: Leiden University Tansley, P, Navaz, H. and M. Roussou (1985). Working with many languages. A handbook for community language teachers. London: School Curriculum Development Committee. Thürmann, E. (1994). Muttersprachliche Bildungsangebote für ausländische Schülerinnen und Schüler im Umbruch. Schulverwaltung NRW 12, 297-299. Tilmatine, M. (ed) (1997). Enseignment des langues d'origine et immigration nordafricaine en Europe: langue maternelle ou langue d'état? Paris: INALCO/CEDREA-CRB. Tinnemans, W. (1991). L'Italianità. De Italiaanse gemeenschap in Nederland. Amsterdam: Het Spinhuis. Tribalat, M. (1995). Faire France. Une enquête sur les immigrés et leurs enfants. Paris: Editions de la Découverte. Uitleg (1990). Invoering Arabische taal en Turkse taal in LBO en MAVO. Regelingen Ministerie van Onderwijs en Wetenschappen 18C, 101-103. Van de Wetering, S. (1990). Het onderwijs in eigen taal en cultuur aan Marokkaanse kinderen in Nederland. PhD thesis, Utrecht: University of Utrecht Van der Avoird, T. (1995). Taalgebruik van en taalverschuiving bij Hindoestanen in Nederland. In: E. Huls and J. KlatterFolmer (eds), Artikelen van de tweede sociolinguïstische conferentie. Delft: Eburon, 21-36. Van der Avoird, T. (1998). Hindi in the United Kingdom and the Netherlands. In: G. Extra and J. Maartens (eds), 195-220. Van der Heijden, M. (1994). Feitelijke (on)mogelijkheden van OET in 's-Hertogenbosch. Master thesis, Tilburg: Tilburg University. Van der Merwe, I. and L. Van Niekerk (1994). Language in South Africa. Distribution and change. Stellenbosch: Dept. of Geography. Van Dijk, N. and J. Lodeweges (1992). Leermiddelen onderwijs in eigen taal (en cultuur). Basisonderwijs en voortgezet onderwijs. Enschede: NICI. Veltman, C. (1983). Language shift in the United States. New York: Mouton. Venes, A. (1994). Zigeunerkinderen in het basisonderwijs. Een studie naar taalvaardigheid. Tilburg: Tilburg University Venes, A. (1995). Onder ons gesproken. Een onderzoek naar taalgebruik van Roma- en Sinti-zigeuners in Nederland. Master thesis, Tilburg: Tilburg University. Verhoeven, L., Narain, G., Extra, G., Konak, Ö. and R. Zerrouk (1995). Toets tweetaligheid. Handleiding, platenboek, leerlingenboeken Turks-Nederlands, Marokkaans Arabisch-Nederlands en Papiamentu-Nederlands. Arnhem: CITO. Verweij, A. (1997). Vaststelling van etnische herkomst in Nederland. De BiZa-methode nader bekeken. Rotterdam: ISEO. White, P. (1990). A question on ethnic group for the census. Findings from the 1989 census test post-enumeration survey. Population Trends 59.

< previous page

page_119

next page >

< previous page

page_ii

next page >

< previous page

page_121

next page > Page 121

Appendix 1 Home Language Survey: Instruction and Questionnaire (translated from Dutch version) (1) Instruction for the Teacher This language survey is being implemented at primary schools in the municipalities of Breda, Eindhoven, Helmond, 'sHertogenbosch, and Tilburg in order to obtain an overview of the languages pupils are exposed to at home. The data to be gathered are very important for the policies pursued at school with respect to instruction in Dutch as a second language and home language instruction to immigrant minority children. The criterion of home language use can also be employed to identify those children in immigrant minority groups that so far have not been able to profit from special government facilities. The language survey is being implemented to make an inventory of the variation in home language use in your year group. Do your pupils use other languages at home instead of or in addition to Dutch? The data gathered will be processed and presented anonymously, with respect to both pupil and school, so that confidentiality will be fully guaranteed. You are requested to present all the questions individually and orally to all the pupils (no group instruction!) QUESTION 11: ANY LANGUAGE OTHER THAN DUTCH? By means of this question, you check whether any language other than Dutch is used in a pupil's home. By 'a language other than Dutch' we do not mean dialects or regional varieties of Dutch (apart from Frisian). We would like you to put questions 12 up to and including 19 only to those children who answer question 11 in the affirmative.

< previous page

page_121

next page >

< previous page

page_122

next page > Page 122

QUESTION 12: NAMES OF LANGUAGES In answer to this question, you write down names of languages rather than names of countries. It is possible that the child will first mention the name of a country where several languages are spoken. If, for example, the pupil says 'Moroccan,' you should check whether in the family Moroccan Arabic or Berber is used. You should bear in mind that pupils from the same country may speak different languages and may not be able to understand one another in these languages. Below, you will find an overview of a number of immigrant minority languages spoken in the Netherlands. ORIGIN Antilles/Aruba China and other countries Great Britain India Indonesia/Moluccan Islands Iran Former Yugoslavia Morocco Spain Surinam Turkey Vietnam Gypsies

EXAMPLES OF LANGUAGES Papiamentu, English, Spanish, ... Chinese, Cantonese, Hakka, ... English, Gaelic (Scottish), Welsh,... Punjabi, Urdu, ... Malay, Javanese, Bahasa (Indonesia), ... Farsi, ... Serbian, Croatian, Slovenian, Albanian,... Moroccan Arabic, Berber,... Spanish, Catalan, Basque, ... Sranan-Tongo, Sarnami-Hindi, Javanese, Hakka, ... Turkish, Kurdish, Armenian, Turoyo, ... Vietnamese, Hmong,... Romanes, Manish, Sinthi, ...

We thank you very much for your cooperation. Should you have any further questions, you can call Tilburg University (Research Group on Language and Minorities).

< previous page

page_122

next page >

< previous page

page_123

next page > Page 123

(2) Questionnaire for the Children

< previous page

page_123

next page >

< previous page

page_124

next page > Page 124

< previous page

page_124

next page >

< previous page

page_125

next page > Page 125

Appendix 2 Questionnaire on Immigrant Minority Language Instruction (IMLI) in Compulsory Education. Please specify for both primary and secondary education separately and successively 1. Goals of IMLI • Are there primarily auxiliary goals for IMLI (e.g., bridging the gap between home and school, and/or contributing to second language acquisition/school success) or are there primarily intrinsic goals for IMLI (i.e., contributing to proficiency in the minority language)? • Are there specified final goals to be reached for proficiency in the immigrant minority language? • If so, what are these goals for comprehension, speaking, reading and writing skills? 2. Target Groups of IMLI • What is the precise definition of the target groups of IMLI? • What criteria are used to identify target groups? (e.g., nationality of person/father/mother, birth-country of person/father/mother, ethnicity based on self-categorization, language)? • How is the concept of 'language' operationalized in the definition and identification of target groups: i.e., Which dimension of the immigrant minority language is taken into account: mother tongue, home language, most regularly used language, standard language in the country of origin? i.e., What is done in case of mismatches between home language and standard language in the country of origin? i.e., How is information on language gathered, e.g., by school, local education authorities, census? • What target groups are legally entitled to IMLI? 3. Policy Instruments for IMLI 3.1. Demand and supply of IMLI • Is there a prior determination of parental and/or pupil demands for IMLI? • How are these demands determined?

< previous page

page_125

next page >

< previous page

page_126

next page > Page 126

• Is there a minimal demand in terms of the number of pupils for a specific language to be offered per class/per school/per municipality? • Are specific immigrant minority groups or languages given priority in IMLI ? • Which languages are offered in IMLI programmes? • What is the absolute and relative degree of participation of different immigrant minority groups in IMLI programs? 3.2 Organization of IMLI • For how many contact hours per week is IMLI given? • When is IMLI taught (within/outside regular school hours?) • For what age groups/classes is IMLI given (all/some)? • Where is IMLI given (the pupils' own school, concentration school, specialized language school)? 3.3 Granting of financial means for IMLI to schools • What formal institution (e.g., ministry, municipality) allocates the financial means for IMLI to schools? • Is allocation based on an open-end budget or a fixed budget? • Is the budget for IMLI an earmarked budget? 3.4 Evaluation of IMLI success. • Is IMLI evaluated in terms of progress in other school subjects (cf. auxiliary goals)? • Is IMLI evaluated in terms of progress in IML proficiency (cf. intrinsic goals)? • Is the degree of IML proficiency expressed in school report marks? 3.5 Teacher training for IMLI • How is teacher training arranged in terms of initial and post-initial programmes? 3.6 Curriculum development for IMLI • Is any co-ordinated action undertaken to develop course materials, dictionaries, tests, etc.? • If so, by what institution(s)? 4. IMLI Policy as Part of a General Language Policy • Do majority language speakers have access to IMLI in primary and/or secondary schools: - de iure? (in principle) - defacto? (in fact)

< previous page

page_126

next page >

< previous page

page_127

next page > Page 127

5. Ideological Context of IMLI Policy • What is the primary rationale for IMLI? (e.g., based on remigration/return perspectives or on the multicultural development of society) • Is there a centralized or decentralized IMLI policy, i.e., a policy focusing on the role of the state/ministry, municipalities, or schools? 6. European Context of IMLI Policy • Is any reference made to European Union directives/actions? • Is any reference made to legislation and/or experiences in other EU countries? 7. Key References Please specify important recent papers, documents and books on IMLI.

< previous page

page_127

next page >

< previous page

page_ii

next page >

< previous page

page_129

next page > Page 129

Appendix 3 Directive 77/486 of the Council of the European Communities on the schooling of children of migrant workers. ANNEX A No L 199/32 Official Journal of the European Communities 6.8.77 (Acts whose publication is not obligatory) COUNCIL COUNCIL DIRECTIVE of 25 July 1977 on the education of the children of migrant workers (77/486/EEC) THE COUNCIL OF THE EUROPEAN COMMUNITIES Having regard to the Treaty establishing the European Economic Community and in particular Article 49 thereof, Having regard to the proposal from the Commission, Having regard to the opinion of the European Parliament(l), Having regard to the opinion of the Economic and Social Committee(2), Whereas in its resolution of 21 January 1974 concerning a social action programme(3), the Council included in its priority actions those designed to improve the conditions of freedom of movement for workers relating in particular to reception and to the education of their children; Whereas in order to permit the integration of such children into the educational environment and the school system of the host State, they should be able to receive suitable tuition including teaching of the language of the host State; Whereas host Member States should also take, in conjunction with the Member States of origin, appropriate measures to promote the teaching of the mother tongue and of the culture of the country of origin of the above mentioned children, with a view principally to facilitating their possible reintegration into the Member State of origin. HAS ADOPTED THIS DIRECTIVE: (1) OJ No C 280, 8.12.1975, p.48 (2) OJ No C 45, 27.2.1976, p.6 (3) OJ No C 13, 12.2.1974, p.1

< previous page

page_129

next page >

< previous page

page_130 Page 130

Article 1 This Directive shall apply to children for whom school attendance is compulsory under the laws of the host State, who are dependents of any worker who is a national of another Member State, where such children are resident in the territory of the Member State in which that national carries on or has carried on an activity as an employed person. Article 2 Member States shall, in accordance with their national circumstances and legal systems, take appropriate measures to ensure that free tuition to facilitate initial reception is offered in their territory to the children referred to in Article 1, including, in particular, the teaching - adapted to the specific needs of such children - of the official language or one of the official languages of the host State. Member States shall take the measures necessary for the training and further training of the teachers who are to provide this tuition. Article 3 Member States shall, in accordance with their national circumstances and legal systems, and in cooperation with States of origin, take appropriate measures to promote, in coordination with normal education, teaching of the mother tongue and culture of the country of origin for the children referred to in Article 1. Article 4 The Member States shall take the necessary measures to comply with this Directive within four years of its notification and shall forthwith inform the Commission thereof. The Member States shall also inform the Commission of all laws, regulations and administrative or other provisions which they adopt in the field governed by this Directive. Article 5 The Member States shall forward to the Commission within five years of the notification of this Directive, and subsequently at regular intervals at the request of the Commission, all relevant information to enable the Commission to report to the Council on the application of this Directive. Article 6 This Directive is addressed to the Member States. Done at Brussels, 25 July 1977. For the Council The President H. SIMONET

< previous page

page_130