How (Unnecessary) Political Tension Leads To Lack Of Linguistic Human Rights In Education: Malsamoj inter versioj

El UEA-vikio
Iri al: navigado, serĉi
e (link fix)
e (2 versioj)
 
(Neniu diferenco)

Nuna versio ekde 10:18, 27 Jun. 2010

How (unnecessary) political tension leads to confusion, controversies, contradictions, inconsistencies and, ultimately, lack of linguistic human rights in education

Invited comments prepared for the Council of Europe conference "Filling the Frames" to mark the 5th anniversary of the entry into force of the Framework Convention on the Protection of National Minorities, Strasbourg, 30-31 October 2003

Tove Skutnabb-Kangas


Introduction

Duncan Wilson’s (hereafter DW) thorough, complex, innovative and very interesting review that I have been invited to comment on, “is a first attempt to collect and critique the opinions of the Advisory Committee on the Framework Convention (ACFC) focussing on rights in, to and through education” (Wilson 2003: 2). It is obviously written from a human rights lawyer’s point of view whereas my comments will look at educational human rights, especially linguistic human rights in education, from a more sociological, sociolinguistic and educational angle. DW concentrates on the articles most relevant for education, namely articles 4, 5, 6, 7, 8, 12, 13 and 14. I intend to focus more narrowly on Article 14. My main (though not only) reason is that this is in many ways the Article that evokes most confusion, contradiction and controversy – and here I certainly agree with DW who has presented fairly bold criticism of both the Article itself and also its interpretation by the Advisory Committee. Ultimately, I claim that the formulation of the article and (therefore?) also some of its interpretations by the Advisory Committee may contribute to depreciation/derogation rather than promotion and protection of educational linguistic human rights, mainly through omission of a principled, research-based stance on (the right to) mother tongue medium education.

The States have, according to the Explanatory note to FCNM, a measure of discretion in the implementation of the objectives (see Note 5 in Wilson 2003). This is one (but only one) of the reasons why the Framework Convention has been seen as pretty toothless by many of us who have criticised it (see, e.g., Alfredsson 2000). In the words of Patrick Thornberry, another leading expert on minority rights:

Despite the presumed good intentions, the provision represents a low point in drafting a minority right; there is just enough substance in the formulation to prevent it becoming completely vacuous (1997: 356-357; see also Thornberry & Gibbons 1997).

But despite this, in the view of DW, “the Advisory Committee must nevertheless identify minimum standards of acceptability in order to monitor the FCNM” (Wilson 2003, First Draft, footnote 3).

The question, then, is to what extent the Advisory Committee have done this? Do they in fact have any leeway to interpret the various Articles, and here especially Article 14, in a more positive way that would agree more with some recent human rights recommendations. Here I am not only thinking of the Hague Recommendations on the Education Rights of National Minorities from the OSCE High Commissioner on National Minorities which DW also refers to. I am also referring to UNESCO’s Education Position Paper (2003a) Education in a multilingual world. Both recommend mother tongue medium education for minorities, for as long as possible .

DW notes that even the Advisory Committee do have some measure of discretion, and that they have indeed used it in relation to several of the other articles (see DW's first paragraph on p. 3), especially article 12. I agree, and join in the congratulations. They have in several opinions suggested constructive measures that seem to extend some positive substantive rights, not only in relation to so-called national minorities but also to some extent in relation to minorities which have been formed on the basis of fairly recent immigration. This is very positive, and in line with what the Human Rights Committee, the Committee on Economic, Social and cultural Rights, the Committee on the Elimination of Racial Discrimination, and the Committee on the Rights of the Child have done. DW comments (Wilson 2003: 7) after stating that these bodies “have often produced clear, imperative recommendations” (something that is to a large extent missing in relation to Article 14 in the Advisory Committee’s work), that this has been “despite at times weak or non-existent provisions in individual treaties on educational rights or minority rights”.

But in relation to Article 14, it seems that the Advisory Committee may not have done their homework thoroughly enough. Their interpretations and recommendations are, unfortunately, full of confusion, contradictions and inconsistencies. And even if they are constantly reminding States about full and effective equality for minority children, many of the recommendations in fact go completely against solid and long-standing research evidence about how best to enable minority children to become high-level bilingual or multilingual through education and do well in school, i.e. have access to good quality education on the basis of equity (and the four A's).

In a charitable reading one might hope that this is because the experts in the Advisory Committee do not know enough about bilingual education research. This also understandable, since they have to cover a very large area of multidisciplinary expertise. In a more pessimistic reading one might suspect that they are merely pragmatically adapting to what they see as possible under present political power relations and realities . In that case I am afraid they are not monitoring state performance but may be participating in what Chomsky calls “manufacturing consent” (Herman & Chomsky 1988).

In the rest of my comments I will discuss two issues in relation to Article 14 and my claims: Issue 1: confusion & controversies; and Issue 2: contradictions and inconsistencies. I will conclude with a few reflections and tentative recommendations.

Issue 1: confusion & controversies

Bilingual education of all kinds is a very specialised and sensitive area of both research and policy-making. Therefore, detailed knowledge of it is a prerequisite for being able to make recommendations. An important complicating issue is that some of the scientifically sound and practically proven principles of how to enable children to become high-level multilingual with the support of the educational system are in fact counter-intuitive: they go against common sense. If indigenous or minority children who speak their mother tongue at home, are to become bilingual, and learn the dominant/majority language well, one might, with a common sense approach, imagine that the principles of early start with and maximum exposure to the dominant language would be good ideas, like they are for learning many other things - practice makes perfect. In fact, sound research shows that the longer indigenous and minority children in a low-status position have their own language as the main medium of teaching, the better they also become in the dominant language, provided, of course, that they have good teaching in it, preferably given by bilingual teachers, just as the Hague Recommendations and the UNESCO guidelines recommend.

I shall give two examples of recent very large-scale longitudinal and methodologically extremely careful studies from the United States, Ramirez et al. (1991) and Thomas & Collier (1997, 2002; see also other references to them in the bibliography). The Ramirez et al.’s 1991 study, with 2352 students, compared three groups of Spanish-speaking minority students. The first group were taught through the medium of English only (but even these students had bilingual teachers and many were taught Spanish as a subject, something that is very unusual in submersion programmes); the second one, early-exit students, had one or two years of Spanish-medium education and were then transferred to English-medium, and the third group, late-exit students, had 4-6 years of Spanish-medium education before being transferred to English-medium. Now the common sense approach would suggest that the ones who started early and had most exposure to English, the English-only students, would have the best results in English, and in mathematics and in educational achievement in general, and that the late-exit students who started late with English-medium education and consequently had least exposure to English, would do worst in English etc.

In fact the results were exactly the opposite. The late-exit students got the best results, and they were the only ones who had a chance to achieve native levels of English later on, whereas the other two groups were, after an initial boost, falling more and more behind, and were judged as probably never being able to catch up to native English-speaking peers in English or general school achievement. The Thomas & Collier study (see bibliography), the largest longitudinal study in the world on the education of minority students, with altogether over 210,000 students, including in-depth studies in both urban and rural settings in the USA, included full immersion programmes in a minority language, dual-medium or two-way bilingual programmes, where both a minority and majority language (mainly Spanish and English) were used as medium of instruction, transitional bilingual education programmes, ESL (English as a second language) programmes, and so-called mainstream (i.e. English-only submersion) programmes. Across all the models, those students who reached the highest levels of both bilingualism and school achievement were the ones where the children's mother tongue was the main medium of education for the most extended period of time. This length of education in the L1 (language 1, first language), was the strongest predictor of both the children's competence and gains in L2, English, and of their school achievement. Thomas & Collier state (2002: 7):

the strongest predictor of L2 student achievement is the amount of formal L1 schooling. The more L1 grade-level schooling, the higher L2 achievement.

The length of mother tongue medium education was in both Ramirez' and Thomas & Collier's studies more important than any other factor (and many were included) in predicting the educational success of bilingual students. It was also much more important than socio-economic status, something extremely vital when reflecting on the socio-economic status discussions and choices in relation to the Roma in the Advisory Committee deliberations and recommendations. The worst results, including high percentages of push-outs ) were with students in regular submersion programmes where the students' mother tongues (L1s) were either not supported at all or where they only had some mother-tongue-as-a-subject instruction. There are dozens of smaller studies from all over the world which show similar results . A typical example would be my own small-scale study among Finnish working class immigrant minorities in metropolitan Stockholm in Sweden (Skutnabb-Kangas 1987). The students in my study were in Finnish-medium classes, and I had Swedish control groups in the parallel classes in the same schools. For their Swedish competence, I used a difficult Swedish language test, of the type where normally middle-class children do better than working class children. After 9 years of mainly Finnish-medium education, and good teaching of Swedish as a second language, these working-class Finnish students got somewhat better results in the Swedish language than the Swedish mainly middle-class control groups. In addition, their Finnish was almost as good as the Finnish of Finnish control groups in Finland.

Table 1. Swedish test results and subjects' own assessment of their Swedish competence

               	TEST RESULT (1-13)	OWN ASSESSMENT(1-5)

M sd M sd Swedish control group 5.42 2.23 4.83 0.26 Finnish co-researchers 5.68 1.86 4.50 0.41

M = mean; sd = standard deviation

Finnish working class immigrant minority youngsters in Sweden, after 9 years of mainly Finnish-medium education; Swedish control group: mainly middle class youngsters in parallel classes in the same schools; Swedish test: decontextualised, CALP-type test where middle-class subjects can be expected to perform better. (Skutnabb-Kangas 1987)

The conclusion from a very thorough research summary for the Maori section of the Ministry of Education in Aotearoa/New Zealand ” (May & Hill 2003: 14) is that English-only submersion programmes “are widely attested as the least effective educationally for minority language students”. Still, it is this kind of submersion education that the Advisory Committee both accepts for many minority students, without protesting, and even recommends for Roma students (see Issue 2, contradictions, below), even to those who speak Romany at home. They follow common sense rather than research results, even if they ought to know that their recommendations are bound to lead to very negative results. They even try to recommend that those minority parents who want to have their children in separate schools, should be dissuaded from this. Knowingly working towards solutions which have been shown to lead to negative results, and not recommending or even advising against solutions which would very likely show positive results, is tantamount to intentionally causing serious mental harm to the children. In fact, this is one of the definitions of genocide in UN’s Genocide Convention :

Article II(b), 'causing serious bodily or mental harm to members of the group'; emphasis added).

None of this is discussed or pointed out in the opinions. There seems to be a complete confusion in relation to what kind of educational models to suggest as the minimum standard. This has been noted by DW too (Wilson 2003: 5), after he has stated that

"Article 14 is the most extreme example of this [specific law, lex specialis, limiting general law, lex generalis which in this case would be the promotion of full and effective equality] for minority rights in education, containing as it does a panoply of geographical, temporal, numerical, demand-based, and more general limitations ('claw-back clauses')". DW then continues: "The Advisory Committee has been particularly hesitant in defining linguistic rights in education, and legal certainty under this article is consequently weak".

Surely the minimum standard cannot be anything that is known through solid research and experience to seriously harm the children? Surely a minimum standard cannot be anything that is known to forcibly assimilate the minority children, as ALL subtractive teaching does. In subtractive language teaching and learning, a new (dominant/majority) language is learned at the expense of the mother tongue. Additive learning and teaching is surely what the Advisory Committee should be recommending for all. In additive language teaching and learning the new language is learned in addition to the mother tongue which continues to be used and developed. For this to happen with minority children, their own language needs to be the main medium of teaching at least during the first 8-9 years of schooling, preferably longer.

Pirjo Janulf (1998) shows in her large longitudinal study that of those Finnish children in Sweden who had been in Swedish-medium submersion classes, with subtractive teaching, not one spoke any Finnish to their own children as adults. The parent generation had no alternatives: Finnish-medium schools and classes did not exist. The subtractive education had as a result that their children were forcibly transferred to the majority group linguistically. This fits another one of the definitions of genocide in UN’s Genocide Convention:

Article II(e), 'forcibly transferring children of the group to another group'.

The political tension leading to the oppression of this kind of reflections as a background to the Advisory Committee’s opinions is indicative of “the controversy surrounding linguistic rights in education” that DW correctly notes (Wilson 2003: 20).

This article is drafted in an extremely complex manner, which well reflects the controversy surrounding linguistic rights in education. The article is full of limitations, or "claw-back clauses". Phrases such as "if there is sufficient demand"; "as far as possible"; "within the framework of their educational systems"; "adequate opportunities for being taught the minority language or receiving instruction in this language" give huge scope for interpretation and endanger the substance of the obligation. Advisory Committee practice under this article appears confused in general, leading to uncertainty as to the balance to be drawn between the various elements of the article, and the criteria for triggering the right laid out in paragraph two. Many of the Advisory Committee’s recommendations under Article 14 are uncommonly weak, often avoiding direct recommendations altogether. (Wilson 2003: 17).

This tension has resulted in that the regulations covering the maintenance of indigenous and minority languages through education are one of the weakest points in all international law covering minority rights. Education mainly through the medium of one’s own language, something that linguistic majorities take for granted, is still not a linguistic human right for most minorities in the world, and the Advisory Committee does nothing to make it so, and adds to the confusion through "avoiding direct recommendations" as noted by DW above.

State reluctance to grant educational linguistic human rights to minorities is based on misplaced and outdated ideologies. These reflect old-fashioned nation-state ideology, a belief that the existence of minorities and their reproduction of themselves as minorities, partly through mother-tongue medium education, necessarily lead to the disintegration of nations states. In fact, it is lack of basic linguistic human rights that contributes to conflict and tension in situations where linguistic hierarchies coincide with political & economic power hierarchies (see, e.g. Eide, Hettne, Stavenhagen, in the bibliography). Granting educational linguistic human rights might be part of a solution, and the Advisory Committee should also contribute to that solution. What these few examples show is the complexity of the question of education of indigenous peoples and minorities. If even advisory committees who are supposedly experts on minority protection show vacillation and uncertainty and confusion, how are educational decision-makers supposed to know what to do? States would do well to heed Robert Phillipson's Recommendation 3 (of 45) in his 2002 book "English-Only Europe? Challenging Language Policy":

3. Each national government must have well-qualified civil servants specializing in language policy, with responsibility for integrating language policies in commerce, culture, education, research, the media, international relations, and in the supranational institutions of the EU system (Phillipson 2002: 180).

Obviously Advisory Committees who work without pay and without reduction of their normal work load, are often doing a fantastic job under difficult circumstances. In a situation already including overwork, one cannot expect them to become experts in all those multidisciplinary fields that their mandates cover. What one can expect, though, is for states to improve their working conditions, and to make access to experts in the various fields easier, an issue that I shall come back to in my concluding remarks. I will discuss issue 2 extremely briefly, only giving a few examples of contradictions and inconsistencies in the Advisory Committee’s opinions, mainly on the basis of DW’s review and its annexes , and using the education of the Roma as one of the examples, to connect better to the next speaker. I see most of the contradictions and inconsistencies as reflections of the confusion that is an inevitable result of the refusal to tackle head-on the complexities and controversies that I have briefly outlined above.

Issue 2: Contradictions and inconsistencies

The Advisory Committee seems unclear on the extent to which ALL education of minorities that does not take place together with children of the dominant group is negative segregation or not. In general, their observations and recommendations in relation to segregation and mainstreaming lack some of the sociologically and educationally necessary distinctions that could be helpful for them in clarifying some of the contradictions and inconsistencies.

The first distinction has to do with whether segregation is used as a goal, or as a means. The goal of some segregation in education has been to disable minority children in the sense that they will never be able to get access to secondary education or higher qualifications because they have been in segregated “special classes” with lower educational standards. This was true in apartheid South Africa’s Bantu education and it is true of some Roma education in Europe (segregation models).

This type of model where also future segregation is a goal, has to be sharply distinguished from minority mother tongue medium classes and schools, which in the nature of things must be separated from dominant language medium classes/schools, but where the goal of this segregation is to enable minority children to become high level bilingual and to achieve in school, to get access to quality education with the same standards as in “mainstream” education. Here segregation is used as a means to equality. These are mother tongue maintenance or language shelter models.

The second, partially overlapping distinction is between permanent or temporary segregation. In much of the negative Roma segregation, just as in the apartheid Bantu education, the educational segregation is a one-way street, both in terms of educational infrastructures and goals. There is no way “back” to “mainstream" education, neither in terms of how the educational infrastructures are organised (quality of curriculum, testing, criteria for access and transfer, etc) or in terms of educational goals/intentions. Once the move has happened, it is permanent in negative segregation. Likewise, the consequences of the negative segregation in education are often permanent marginalisation on the labour market and in social and political participation in society in general.

In contrast, in the positive “segregation" models, the segregation is temporary and necessary. The goal (which is reached) is to enable the students to develop linguistic, psychological, social and educational qualities/characteristics and competencies which make it possible for them to integrate themselves in the larger society at some point in both their educational career (at the earliest after the primary cycle, often only in upper secondary or tertiary studies) and, especially, after school. Thus the temporary segregation leads to later positive integration, i.e. social, educational, labour market oriented and political incorporation, but without assimilation.

The third potentially helpful distinction is between physical versus psychological and linguistic segregation or integration. Temporary physical segregation is often necessary for minority children in order to reach later psychological and linguistic integration (as opposed to marginalisation and forced assimilation). It is perfectly possible to “integrate” a minority child in preschool and school, in a “mainstream” classroom with dominant group children, physically, yes. But if the child does not speak the dominant language, the medium of instruction, as her mother tongue, her chances of accessing knowledge at the same level as the majority children, are seriously reduced, as one can see from the studies reviewed above. Mostly her own mother tongue does not have a chance of developing in this “integrated” class. A few weekly hours of mother tongue teaching as a subject may be psychologically positive for identity but are completely insufficient for developing the mother tongue up to a high linguistic and cognitive level. This leaves the child without a firm basis for learning the dominant language too, and thus the child’s bilingualism is only at a surface level, sufficient for everyday interactions at face-to-face level when discussing concrete and contextualised issues, but completely insufficient for academic achievement and decontextualised problem solving already towards the end of primary education. And these children are probably never going to catch up with dominant language children, as research also shows. Those children who succeed, do it despite the school, not because of the way their education is organised. Psychologically the children are mostly at a disadvantage and often psychologically and socially segregated in the classroom, also because of discrimination and often low social status. It is often only their bodies which are integrated while their minds, souls and hearts can be alone and remain segregated, as dozens of works of fiction have described recently. Thus early physical integration often leads to later physical, linguistic, psychological, social, socio-economic and political segregation.

Again, in contrast, early voluntary and temporary physical segregation from dominant-language peers in mother tongue medium maintenance classes for minority children enable them to develop their mother tongues up to a high formal level, to learn the dominant language thoroughly (as well as or even better than the dominant language-speaking peers, as my study showed), to learn the content of education because they understand it, and in general develop all their capacities if the education is of high standards as it should be. The children are in fact perfectly integrated in their classrooms where they understand the instruction and can develop and grow and show what they know. Thus this initial physical segregation from the dominant society children enables them to integrate themselves later on in the larger society, both physically and psychologically, linguistically, socially, socio-economically and politically. One could add a fourth distinction which is implicit in the three others, namely segregation from or integration into society as a whole versus school. Segregation from dominant group children in school (if it is in a maintenance model, as opposed to a segregation model) can lead to integration into society – and this is what should be the goal.

If we now look at the observations and recommendations of the Advisory Committee, we can see that most of the contradictions and inconsistencies are due to the Committee not having been given the time to clarify these basic distinctions. DW quotes the Committee on Economic, Social and Cultural Rights, on Bulgaria (ICESCR E/2000/22 (1999) 46; Wilson 2003, Annex 2, Note 129):

239. The State party should continue its efforts to integrate ethnic minorities into society, and undertake measures to provide the opportunity for such minorities to be educated in their own languages (emphases added).

In the quote, the fourth distinction has clearly been made. What the quote suggests is that minorities can have their education through the medium of their own language (which necessarily means physical segregation in school from dominant language speakers during most lessons) and still be integrated into the larger society.

How has the Advisory Committee handled the question of integration versus segregation? As DW notes (Wilson 2003: 18), it has consistently used the formulation “placing children in separate special classes should take place only when it is absolutely necessary and always on the basis of consistent, objective and comprehensive tests”. The Advisory Committee does NOT say exactly what these “separate special classes” are. Provided they mean a negative segregation model described above, the formulation is obviously a good attempt to protect children from discrimination. The problem is that mother tongue maintenance classes are also necessarily “separate”. And mother tongue maintenance classes can also be labelled “special”, if one does what the Advisory Committee does, namely call majority-medium classes and schools “ordinary education” (e.g. Wilson 2003: 4, and Note 18) or “mainstream” or “regular” (e.g. ibid.: 38). For a minority child to be placed in a “separate special” maintenance model class, no testing is needed – it should be enough that the child has a minority mother tongue. In several opinions, the Advisory Committee urges the state to place the minority children (mostly Roma, but also some others) in “regular” classes where they are “integrated” with (i.e. physically integrated with but often psychologically segregated from) dominant group children. At the same time, the Committee urges the State to “ensure [that] adequate opportunities exist to be taught the Roma language or to receive instruction in this language” (e.g. opinions on Romania, Croatia and Slovakia, see Wilson 2003: 17-18 and pp. 32, 41; they also see “separate classes” as risky for integration – e.g. in the opinion on Sweden). But you cannot teach minority children through the medium of a minority language, in a classroom where children from the linguistic majority are also present, unless these are also to become bilingual – and there is no indication of this in the Advisory Committee’s opinions. This is a serious contradiction.

If there are both minority and majority children in the classroom, and part or most of the teaching is through the medium of the minority language, it should be a properly planned “two-way bilingual model” where all children are supposed to become bilingual (see Dolson & Lindholm 1995, Lindholm 1997, Lindholm-Leary 2001 for these). Maybe the Advisory Committee should start looking at both two-way bilingual models and also immersion models – both are very popular in Canada and the USA, immersion also in Catalunya, the Basque Country and Finland. Or they might suggest models built on the principles in the special European Union Schools model (see articles in Skutnabb-Kangas, ed. 1995, for these).

As DW also notes (e.g. Wilson 2003: 17), the Advisory Committee is completely inconsistent in the length of time that it recommends for mother tongue medium education for various groups and countries. Especially in the case of the Roma, the Committee is actively preventing ANY mother tongue medium education, through their insistence on the children being in “regular” classes, together with dominant language speakers. Likewise, in several cases, the Committee does not clearly protest when there are clear violations of the Convention, as when Roma children have no or very little instruction in their language as a subject and absolutely no teaching through the medium of their own language (see Wilson 2003: 20 for Romania; in the opinion on the Slovak Republic, there is no clear protest against no education being given through the medium of Romany even when the Slovak Constitution “guarantees the rights of Slovak citizens belonging to national minorities to receive education in their mother tongue”; ibid.: 47). For others, the Committee do not think that 4 years (e.g. opinion on Austria, Wilson 2003: 46) or primary and lower secondary education (e.g. Estonia, Sweden, ibid.: 46) is enough. For still others (e.g. in opinions on Albania, Germany, Romania and the Slovak Republic, Wilson 2003: 36-37) they consider minority worries about lack of university education through the medium of their own languages as legitimate. Inconsistency. Likewise, the Advisory Committee considers, according to DW (Wilson 2003: 39), on the one hand, that “specific classes devoted to one national minority as such … risk placing the children concerned at a disadvantage and harming the implementation of Article 12” (about the Roma), whereas there are no worries about placing other minorities in mother tongue medium education up to university level. Contradiction and inconsistency. In many – maybe most – opinions, the Advisory Committee seems to listen to parents and minority organisations, using their opinions and wishes to promote legal protection of minority languages and minority language medium education. In the case of the Roma and Travellers (e.g. Sweden and the UK, Wilson 2003: 39), the Committee wants the State to try, directly or indirectly, to dissuade the Roma and/or Traveller children’s parents from wanting separate schools for their children (schools that, as we have seen, are the only ones that could use the minority language as a medium of instruction). Again, contradiction and inconsistency.

The opinions in relation to Article 14, and especially in relation to mother tongue medium education, thus seem to be contextualised, recommending different solutions to different countries and minorities. This could be positive, if it was done on the basis of pricipled solutions, following best practices and theories from solid research. Instead, the recommendations seem to be contextualised on the basis of political power relations, and with too little background knowledge about educational models that work and consequences of models that do not work.

To conclude

As I hope to have shown, issues about linguistic human rights in education are unusually complex. Educational linguistic human rights, especially an unconditional rights to mother tongue medium education, are in many indigenous peoples' and minorities' views one of the two or three most important issues for their ability to reproduce themselves as indigenous peoples/minorities, the other ones being minimally cultural autonomy for minorities and self-determination for indigenous peoples, in addition to the land rights question for indigenous peoples. This is something that states should acknowledge. But in many if not most cases this does not happen, or happens only partially and reluctantly.

The pessimistic (but realistic) assessment by Rodolfo Stavenhagen (1995: 76-77; see also Stavenhagen 1996) is:

Too often, policies of national integration, of national cultural development, actually imply a policy of ethnocide, that is, the wilful destruction of cultural groups ... The cultural development of peoples, whether minorities or majorities, must be considered within the framework of the right of peoples to self-determination, which by accepted international standards is the fundamental human right, in the absence of which all other human rights cannot really be enjoyed ... governments fear that if minority peoples hold the right to self-determination in the sense of a right to full political independence, then existing States might break up ... State interests thus are still more powerful at the present time than the human rights of peoples (emphasis added).

It is imperative for the Advisory Committee to see its work in this broader light. This implies that it must, within "the discretion in the implementation of the objectives" that states have, and the leeway they may have to interpret the various Articles, and here especially Article 14, "identify minimum standards of acceptability in order to monitor the FCNM". The minimum standards need to build on solid research knowledge about how the goals of the Framework Convention can be concretised, in the way Duncan Wilson has described this, for instance within the 4 A's scheme, with Availability, Accessibility, Acceptability, and Adaptability. More concretisation, of both the differences between the various educational models and of principles derived from successful education, are shown in my Tables 2, 3 and 4 , in the Annex.

The Advisory Committee is obviously walking a very tight rope, having to balance between what minority children need and have a right to in education, and what states are willing to grant to them. It has already courageously broadened the scope of what can be suggested, for instance in its attempts to include immigrant and refugee minorities, rather than just long-standing national minorities, in its recommendations. But as a watchdog, they need to do still more. It is imperative to grasp the hot potato of (the right to) mother tongue medium education, where very little has happened since the famous Belgian Linguistic Case .

Many people react to hearing "genocide" in this context by an emotional blackout of some kind so that they are unable to discuss the substantive issue: does most of today's minority education in fact fulfill the requirements in the two UN Genocide Convention genocide definitions or not? What we need is a discussion of the issues in a calm way, rather than simply censoring/forbidding a discussion because of the emotional reaction. It is extremely clear that no teachers or principals of schools (or, for that matter, members of any Advisory Committee) plan to contribute to linguistic genocide with their actions and decisions in school. Today's linguistic genocide is no longer a question of beating up children or putting them in jail for speaking their own language - this happens in very few countries today (but it does nevertheless happen - see, e.g., Skutnabb-Kangas & Bucak 1994, Skutnabb-Kangas 2002). Today's linguistic genocide does not have that kind of individual actors - the agency is embedded in structures which accomplish what the Genocide Convention's Final Draft in its Article 3(1) defined as linguistic genocide:

prohibiting the use of the language of the group in daily intercourse or in schools, or the printing and circulation of publications in the language of the group

Without an unconditional right to mother tongue medium education in schools, the use of the language of the [minority] group is factually prohibited - children cannot use their language during the lessons as a medium of teaching and learning. The utter lack of clarity of any Advisory Committee position on linguistic human rights in education may well make the Advisory Committee guilty by omission for failing to protect the linguistic identity and the best interests of the child, as they are supposed to do. If States Parties are ignorant of their own agency here, it is the duty of the Advisory Committee to inform them.

For something to fulfill the genocide definitions, there has to be an intention. It is in my view difficult for states and school authorities to plead not guilty and claim that they had no negative intentions. The knowledge about how harmful most dominant-language-medium education has been for indigenous and minority children and how it has forcibly tried, and often succeeded in, assimilating them linguistically and culturally (i.e. it has forcibly transferred children from their own group to the dominant group), has existed widely among church and school authorities for almost a century and a half; research results have been there almost a century, and even modern research dates as a minimum from fifty years ago, from the time of UNESCO's classic book The use of vernacular languages in education (1953) which has just been replaced as UNESCO guidelines by the book mentioned above, Education in a multilingual world (2003). The knowledge and research is there; now the Advisory Committee needs to use it to persuade reluctant states to grant and implement those basic linguistic human rights in education in relation to minority children which dominant language speakers take for granted for themselves.

It is also difficult for ordinary minority parents to know enough about educational and linguistic research to make demands for rights which are counter-intuitive. Many parents may be fooled by dominant group (or even sometimes their own) school authority representatives into believing in the common sense based solutions, or into thinking that they must choose between the two languages: either mother tongue and identity or the dominant language and a job. These unnecessary either-or solutions, instead of the additive both-and, are all to common. When parents have enough solid research-based knowledge about the long-term consequences of various choices, and understand the theoretical basis behind additive teaching, most parents choose mother tongue medium based solutions. This can be seen very clearly among those (both immigrant and national - see articles in Peura & Skutnabb-Kangas, eds. 1994) minorities and indigenous peoples who now, like the Saami in Norway and Finland, have demanded and got some rights (Magga 1995) and have their own schools (see Aikio-Puoskari & Skutnabb-Kangas 2002). It should be a human right for parents to have enough background knowledge to make informed choices; and once they have made their choices, the alternatives should be there. The Advisory Committee can support this.

All this means, in terms of recommendations, that the Advisory Committee needs to have proper opportunities, also in terms of time and remuneration, to consider the issues they deal with, within thematic blocs, where they have the chance to get the basic expert research knowledge they need as a basis, in each multidisciplinary area, before they need to make recommendations. They also need the time for dispassionate discussion and consideration of their recommendations, including an assessment of the impact that various alternatives have on minority children's educational rights and achievement. This is the only way to bring more clarity and consistency in the recommendations, rather than the vacillation, uncertainty and confusion that, despite good intentions and hard work, is apparent in some of the recommendations of the Advisory Committee. Starting with the theme of educational linguistic human rights might be an appropriate urgent first thematic block.

Epilogue - ethics and research

I would like to remind those who might find some parts of my paper somewhat unsettling, of Edward Said's words about the role of intellectuals. Said, himself one of the greatest intellectuals of our times, passed away a few weeks ago:

An intellectual is 'neither a pacifier nor a consensus-builder, but someone whose whole being is staked on a critical sense, a sense of being unwilling to accept easy formulas, or ready-made clichés, or the smooth, ever-so-accommodating confirmations of what the powerful or conventional have to say, and what they do. Not just passively unwilling, but actively willing to say so in public.' (Said 1994: 17).

'The intellectual is an individual endowed with a faculty for representing, embodying, articulating a message, a view, an attitude, philosophy or opinion to, as well as for, a public. And this role has an edge to it, and cannot be played without a sense of being someone whose place it is publicly to raise embarrassing questions, to confront orthodoxy and dogma (rather than to produce them), to be someone who cannot be easily co-opted by governments or corporations ... Least of all should an intellectual be there to make his/her audience feel good: the whole point is to be embarrassing, contrary, even unpleasant' (ibid.: 9-10).

'There is no question in my mind that the intellectual belongs on the same side with the weak and unrepresented.' (ibid.: 27). These are for Said 'human beings considered to have subaltern status, minorities, small peoples and states, inferior cultures and races' (ibid.: 27). Said's 'modern intellectual's role' is truly 'that of disputing the prevailing norms' (ibid.: 27).


References:

Aikio-Puoskari, Ulla & Skutnabb-Kangas, Tove (in press). When few under 50 speak the language as a first language: linguistic (human) rights and linguistic challenges for endangered Saami languages. In Helander, Nils Øyvind (ed.). xx.. Guovdageaidnu/Kautokeino: Sámi Instituhhta/Nordic Saami Institute.

Alfredsson, Gudmundur (2000). A Frame with an Incomplete Painting. Comparison of the Framework Convention for the Protection of National Minorities with international standards and monitoring procedures. International Journal on Minority and Group Rights 7:41-304.

Baker, Colin (1993). Foundations of Bilingual Education and Bilingualism. Clevedon & Philadelphia: Multilingual Matters.

Baker, Colin & Prys Jones, Sylvia (1998). Encyclopedia of Bilingualism and Bilingual Education. Clevedon/Philadelphia: Multilingual Matters.

Collier, Virginia P. (1989). How long? A synthesis of research on academic achievement in a second language, TESOL Quarterly 23, 509-531.

Collier, Virginia P. (1987). Age and rate of acquisition of second language for academic purposes, TESOL Quarterly 21, 617-641.

Collier, Virginia P. & Thomas, Wayne P. (2002). Reforming Education Policies for English Learners Means Better Schools for All. The State Educational Standards 3:1, 30-36.

Cummins, Jim (1989). Empowering Minority Students, Sacramento, California Association for Bilingual Education.

Cummins, Jim (1991). The development of bilingual proficiency from home to school: A longitudinal study of Portuguese-speaking children. Journal of Education, 173(2), 85-98.

Cummins, Jim (1992a). Bilingual Education and English Immersion: The Ramírez Report in Theoretical Perspective. Bilingual Research Journal 16:1-2, 91-104.

Cummins, Jim (1994a). The discourse of disinformation: the debate on bilingual education and language rights in the United States. In Skutnabb-Kangas & Phillipson (eds), 159-177.

Cummins, Jim (1995). Reflections on the European Schools Model in relation to French Immersion Programs in Canada, in Skutnabb-Kangas (ed.), 159-168.

Cummins, Jim (1996). Negotiating Identities: Education for Empowerment in a Diverse Society. Ontario, California: California Association for Bilingual Education.

Cummins, Jim (2000). Language, Power, and Pedagogy: Bilingual Children in the Crossfire. Clevedon, UK: Multilingual Matters.

Cummins, Jim & Corson, David (eds) (1997). Bilingual Education. Volume 5. Encyclopedia of Language and Education. Dordrecht, Boston & London: Kluwer Academic Publishers.

Dolson, David & Lindholm, Kathryn (1995). World class education for children in California: A comparison of the two-way bilingual immersion and European School model. In Skutnabb-Kangas, Tove (ed.). Multilingualism for All. Lisse: Swets & Zeitlinger, 69-102.

Eide, Asbjørn (1990). Possible ways and means of facilitating the peaceful and constructive solution of problems involving minorities. Progress report submitted to Sub-Commission on Prevention of Discrimination and Protection of Minorities at its 42nd session (E/CN.4/Sub.2/1990/46). New York: United Nations.

Eide, Asbjørn (1991). Preliminary report submitted to UN Sub-Commission on Prevention of Discrimination and Protection of Minorities at its 43rd session (E/CN.4/Sub.2/1991/43). New York: United Nations.

Eide, Asbjørn (1993). Possible ways and means to facilitate the peaceful and constructive solution of problems involving racial minorities. E/CN.4/Sub.2/1993/34 and Add.1-4. New York: United Nations.

Eide, Asbjørn (1994). New approaches to minority protection, London: Minority Rights Group International.

The Hague Recommendations Regarding the Education Rights of National Minorities & Explanatory Note, October 1996, The Hague; for the use of the OSCE High Commissioner on National Minorities, Max van der Stoel. Xx web

Herman, Edward S., and Chomsky, Noam (1988). Manufacturing Consent: the Political Economy of the Mass Media. New York: Pantheon.

Hettne, Björn (1987). Etniska konflikter och internationella relationer [Ethnic conflicts and international relations]. Rapport 6 från DEIFO. Stockholm: DEIFO (Delegationen för invandrarforskning).

Hettne, Björn (1990). Development Theory and the Three Worlds. Harlow: Longman.

Huss, Leena (1999). Reversing Language Shift in the Far North. Linguistic Revitalization in Scandinavia and Finland. Acta Universitatis Upsaliensis. Studia Uralica Upsaliensia 31. Uppsala: Uppsala University.

Huss, Leena, Camilleri, Antoinette & King, Kendall (eds) (2003). Transcending Monolingualism: Linguistic Revitalisation in Education. Series Multilingualism and linguistic diversity. Lisse: Swets & Zeitlinger.

Lindholm, Kathryn J. (1997). Two-Way Bilingual Education Programs in the United States. In Cummins, Jim & Corson, David (eds) (1997). Bilingual Education. Volume 5. Encyclopedia of Language and Education. Dordrecht, Boston & London: Kluwer Academic Publishers, 271-280.

Lindholm-Leary, Kathryn J. (2001). Dual Language Education. Clevedon, Buffalo, Toronto & Sydney: Multilingual Matters.

Magga, Ole Henrik (1994). The Sámi Language Act. In Skutnabb-Kangas, Tove and Phillipson, Robert (eds.), Linguistic Human Rights. Overcoming linguistic discrimination. Berlin and New York: Mouton de Gruyter, 219-233.

May, Stephen & Hill, Richard (2003). Bilingual/Immersion Education: Indicators of Good Practice. Milestone Report 2. Hamilton: Wilf Malcolm Institute of Educational Research, School of Education, University of Waikato.

May, Stephen, Hill, Richard & Tiakiwai Sarah (2003). Bilingual/Immersion education: Indicators of Good Practice. Milestone Report 1.

Peura, Markku & Skutnabb-Kangas, Tove (eds.) (1994). "Man kan vara tvåländare också". Den sverigefinska minoritetens väg från tystnad till kamp. (You can be bicountrial too. The road of the Sweden-Finnish minority from silence to struggle). Sverigefinländarnas Arkiv, Stockholm.

Phillipson, Robert (2003). English-Only Europe? Challenging Language Policy. London: Routledge.

Ramirez, J. David (1992). Executive Summary. Bilingual Research Journal, 16(1&2), 1-62.

Ramirez, J. David, Pasta, David J., Yuen, Sandra D., Billings, David K. & Ramey, Dena R. (1991). Longitudinal study of structured immersion strategy, early-exit, and late-exit bilingual education programs for language minority children (Vols. 1-2). San Mateo, CA: Aguirre International.

Ramirez, J. David., Yuen, Sandra D. & Ramey, Dena R. (1991). Executive Summary: Final report: Longitudinal study of structured English immersion strategy, early-exit and late-exit transitional bilingual education programs for language-minority children, Submitted to the U. S. Department of Education. San Mateo, CA: Aguirre International.

Said, W. Edward (1994). Representations of the intellectual. London: Vintage.

Sieghart, Paul (1983). The international law of human rights, Oxford: Oxford University Press.

Skutnabb-Kangas, Tove (1984). Bilingualism or not ﷓ the education of minorities. Clevedon, Avon: Multilingual Matters.

Skutnabb-Kangas, Tove (1987). Are the Finns in Sweden an Ethnic Minority? Finnish Parents Talk about Finland and Sweden. Research Project "The Education of the Finnish Minority in Sweden", Working Paper nr 1. Roskilde: Roskilde University Centre, Institute VI.

Skutnabb-Kangas, Tove (2000). Linguistic genocide in education – or worldwide diversity and human rights? Mahwah, New Jersey & London: Lawrence Erlbaum Associates.

Skutnabb-Kangas, Tove (2002). Linguistic human rights in education and Turkey - some international comparisons. Invited plenary paper at the International Conference on Kurds, the European Union and Turkey, Copenhagen, Denmark, 14th October 2002. At http://www.kurds.dk/english/2000/news115.html and http://www.kurdistan.nu/english/tove.htm. In Danish: Sproglige rettigheder i undervisning og Tyrkiet - nogle internationale sammenligninger. I Dokumenter fra International Konference om Kurdere, EU og Tyrkiet, 14-15 oktober 2002. København: KOMKAR (Den Kurdiske Forening) & AIF (Arbejderbevægelsens Internationale Forum), 20-41.

Skutnabb-Kangas, Tove & Bucak, Sertaç (1994). Killing a mother tongue ﷓ how the Kurds are deprived of linguistic human rights. In Skutnabb-Kangas, Tove & Phillipson, Robert (eds.), in collaboration with Mart Rannut. Linguistic Human Rights. Overcoming Linguistic Discrimination. Contributions to the Sociology of Language 67. Berlin & New York: Mouton de Gruyter, 347-370.

Skutnabb-Kangas, Tove & García, Ofelia (1995). Multilingualism for All - General Principles? In Tove Skutnabb-Kangas (Ed.). Multilingualism for All, Lisse: Swets & Zeitlinger, Series European Studies on Multilingualism, 221-256.

Skutnabb-Kangas, Tove (ed.) (1995). Multilingualism for All. Series European Studies on Multilingualism. Lisse, The Netherlands: Swets & Zeitlinger.

Stavenhagen, Rodolfo (1990). The Ethnic Question. Conflicts, Development, and Human Rights, Tokyo: United Nations University Press.

Stavenhagen, Rodolfo (1995). Cultural rights and universal human rights. In Eide, Asbjørn, Krause, Catarina & Rosas, Allan (eds.) 1995. Economic, Social and Cultural Rights. A Textbook. Dordrecht, Boston & London: Martinus Nijhoff Publishers, 63-77.

Stavenhagen, Rodolfo (1996). Self-Determination: Right or Demon? In Clark, Don & Williamson, Robert (eds). Self-Determination: International Perspectives. London: The Macmillan Press, 1-11.

Thomas, Wayne P. and Collier, Virginia (2002). A National Study of School Effectiveness for Language Minority Students' Long-Term Academic Achievement Report: Project 1:1. Fairfax, Virginia: VREDE, George Mason University. www.crede.ucsc.edu.

Thomas, Wayne P. and Collier, Virginia P. (2002). A National Study of School Effectiveness for Language Minority Students' Long Term Academic Achievement. George Mason University, CREDE (Center for Research on Education, Diversity & Excellence). http://www.crede.ucsc.edu/research/llaa/1.1_final.html.

Thornberry, Patrick (1997). Minority Rights. In Academy of European Law (ed.). Collected Courses of the Academy of European Law. Volume VI, Book 2, 307-390. The Netherlands: Kluwer Law International.

Thornberry, Patrick & Gibbons, Dianna (1997). Education and Minority Rights: A Short Survey of International Standards. International Journal on Minority and Group Rights. Special Issue on the Education Rights of National Minorities 4:2, 1996/1997, 115-152. UNESCO 2003a. Education in a multilingual world. UNESCO Education Position Paper. Paris: UNESCO. http://unesdoc.unesco.org/images/0012/001297/129728e.pdf UNESCO 2003b. Language Vitality and Endangerment. UNESCO Intangible Cultural Heritage Unit’s Ad Hoc Expert Group on Endangered Languages. Approved 31 March 2003 by the Participants of the at International Expert Meeting on UNESCO Programme Safeguarding of Endangered Languages, UNESCO, Paris-Fontenoy,10-12 March 2003. In press. UNESCO 2003c. Recommendations for Action Plan. International Expert Meeting on UNESCO Programme Safeguarding of Endangered Languages, UNESCO, Paris-Fontenoy,10-12 March 2003. In press.

van der Stoel, Max (1999). Report on the Linguistic Rights of Persons Belonging to National Minorities in the OSCE area. + Annex. Replies from OSCE Participating States. The Hague: OSCE High Commissioner on National Minorities.

Some definitions:

An immersion programme is a programme where linguistic majority children with a high-status mother tongue voluntarily choose (among existing alternatives) to be instructed through the medium of a foreign (minority) language, in classes with majority children with the same mother tongue only (but see below) where the teacher is bilingual so that the children can at the beginning use their own language, and where their mother tongue is in no danger of not developing or of being replaced by the language of instruction - an ADDITIVE language learning situation. (Skutnabb-Kangas 2000, Definition Box 8.6)

A language maintenance programme or language shelter programme is a programme where linguistic minority children (often with a low-status mother tongue) voluntary choose (among existing alternatives) to be instructed through the medium of their own mother tongue, in classes with minority children with the same mother tongue only, where the teacher is bilingual (almost always in the case of immigrant and refugee minorities and indigenous peoples; not always (but often) to the same extent in the case of national minorities) and where they get good teaching in the majority language as a second/foreign language, also given by a bilingual teacher. (Skutnabb-Kangas 2000, Definition Box 8.5)

A segregation model is a programme where linguistic minority children with a low-status mother tongue are forced to accept instruction through the medium of their own mother tongue in classes with minority children (with the same mother tongue) only, where the teacher may be monolingual or bilingual but is often poorly trained, where the class/school has poorer facilities and fewer resources than classes/schools for dominant group children (see Info Box 6.9), and where the teaching of the dominant language as a second/foreign language is poor or non-existent. (Skutnabb-Kangas 2000, Definition Box 8.3)

A submersion or sink-or-swim model is a programme where linguistic minority children with a low-status mother tongue are forced to accept instruction through the medium of a foreign majority/official language with high status, in classes where some children are native speakers of the language of instruction, where the teacher does not understand the mother tongue of the minority children, and where the majority language constitutes a threat to the minority children's mother tongue (MT), which runs the risk of being displaced or replaced (MT is not being learned (at a high level); MT is 'forgotten'; MT does not develop because the children are forbidden to use it or are made to feel ashamed of it) - a subtractive language learning situation. (Skutnabb-Kangas 2000, Definition Box 8.2)

[In] subtractive language learning, a new (dominant/majority) language is learned at the cost of the mother tongue which is displaced, with a diglossic situation as a consequence, and is sometimes replaced. The person's total linguistic repertoire does not show (much) growth as a result of the learning. The concept originally comes from Wallace Lambert (1975).

[In] additive language learning - a new language is learned in addition to the mother tongue which continues to be used and developed. The person's total linguistic repertoire is extended (from Skutnabb-Kangas 2000, Definition Box 2.2).

Two﷓way bilingual (dual language) (also called 'dual immersion' in the USA - see 8.3.1 for why this is a misnomer) models are models with approximately 50% majority students and 50% minority students (with the same mother tongue) who voluntarily choose to be instructed by a 'completely' bilingual teacher, initially mainly through the medium of a minority language (the 90%-10% model) or through the medium of both languages (the 50%-50% model), with the dominant language taught as a subject (at the beginning separately to both groups: mother tongue English to native English-speakers and English as a second language to minority language speakers in the USA). The percentage of instruction in the dominant language increases in all 90-10 models, in some to 40-50% or even 60% by grade 6, whereas it stays the same in the 50-50 model (something that actually would place them in the transitional models category). Two-way models thus combine in one classroom a maintenance model for minorities (especially in the 90-10 model) and an immersion model for the majority while maximizing peer-group contact in the other language for both groups. (Skutnabb-Kangas 2000, Definition Box 8.37

A transitional programme is a programme where linguistic minority children with a low-status mother tongue are initially instructed through the medium of their mother tongue for a few years and where their mother tongue is taught as though it has no intrinsic value, only an instrumental value. It is used only in order for the children to learn the majority language better, and so as to give them some subject matter (content) knowledge while they are learning the majority language. In early-exit programmes, children are transferred to a majority language medium programme as soon as they can function to some extent in the majority language orally, in most cases after one or two, at the most three, years. In late-exit programmes they may continue to have at least some of their education, sometimes up to half of it, in L1 up to the 5th or 6th grade, and sometimes the mother tongue may be taught as a subject even after that. 4Skutnabb-Kangas 2000, Definition Box 8.3)


Annex

Table 2. 8 Principles followed in successful educational programmes which lead to good results in terms of high levels of bi- or multilingualism, a fair chance of success in school achievement, and positive multilingual/multicultural identities and attitudes. 1. Support (= use as the main medium of education, at least during the first 8 years) that language (of the two that the child is supposed to become bilingual in initially) which is least likely to develop up to a high formal level. This is for all minority children their own mother tongue. For majority children, it should be a minority language. (The European Schools do not follow this principle completely, because they teach also majority children initially through the medium of their mother tongues, e.g. the the Italian-speaking children in the European School in Italy are initially taught through the medium of Italian, instead of a minority language).

2. In most experiments, the children are initially grouped together according to their L1. Mixed groups are not positive initially, and certainly not in cognitively demanding decontextualised subjects. (Spanish-English Two-way programmes in the U.S.A. are an exception: they have mixed in the same class 50% minority, 50% majority children. All are initially taught through the medium of the minority language, later through both. This may be a relevant factor in accounting for the Spanish-speaking children's sometimes relatively less impressive gains in both languages, compared to English-speaking children in the same programmes. The mere presence of majority language children in the same classroom may be too overwhelming for minority children, despite the minority language being the medium of education).

3. All children are to become high level bilinguals, not only minority children. This seems to be especially important in contexts where majority and minority children are in the same classes.

4. All children have to be equalized vis-a-vis the status of their mother tongues and their knowledge of the language of instruction. Nice phrases about the worth of everybody's mother tongue, the value of interculturalism, etc, serve little purpose, unless they are followed up in how the schools are organised.

There has to be equality in the demands made on the children's and the teachers' competencies in the different languages involved, so that the same demands are made on everybody. Both minority and majority children and teachers must be or become bi- or multilingual.

There has to be equality in the role that the languages are accorded on the schedules and in higher education, in testing and evaluation, in marks given for the languages, in the physical environment (signs, forms, letters, the school's languages of administration, the languages of meetings, assemblies, etc), in the status and salaries of the teachers, in their working conditions, career patterns, etc.

It is possible to equalize the children vis-a-vis their knowledge of the language of instruction in several different ways:

A. All children know the language of instruction (maintenance programmes, European Schools initially);

B. No children know the language of instruction or everybody is in the process of learning it (immersion programmes, European Schools in certain subjects in a later phase);

C. All children alternate between 'knowing' and 'not knowing' the language of instruction (two-way programmes in a late phase; alternate-days programmes (50% minority and 50% majority children, the medium of education alternates daily).

5. All teachers have to be bi- or multilingual. Thus they can be good models for the children, and support them in language learning, through comparing and contrasting, and being metalinguistically aware. Every child in a school has to be able to talk to an adult with the same native language.

This demand is often experienced as extremely threatening by majority group teachers, many of whom are not bilingual. Of course all minority group teachers are not high level bilinguals either. But it is often less important that the teacher's competence in a majority language is at top level, for instance in relation to pronunciation, because all children have ample opportunities to hear and read native models of a majority language outside the school anyway, whereas many of them do NOT have the same opportunities to hear/read native minority language models. A high level of competence in a minority language is thus more important for a teacher than a high level of competence in a majority language.

6. Foreign languages should be taught through the medium of the children's mother tongue and/or by teachers who know the children's mother tongue. No teaching in foreign languages as subjects should be given through the medium of other foreign languages (for instance, Turkish children in Germany should not be taught English through the medium of German, but via Turkish).

7. All children must study both L1 and L2 as compulsory subjects through grades 1-12. Both languages have to be studied in ways which reflect what they are for the children: mother tongues, or second or foreign languages. Many minority children are forced to study a majority language, their L2, as if it was their L1.

8. Both languages have to be used as media of education in some phase of the children's education, but the progression in how and how much each is used seems to vary for minority and majority children.

For MAJORITY CHILDREN the mother tongue must function as the medium of education at least in some cognitively demanding, decontextualized subjects, at least in grades 8-12, possibly even earlier.

But MAJORITY CHILDREN can be taught through the medium of L2 at least in some (or even all or almost all) cognitively less demanding context-embedded subjects from the very beginning. L2 can also be the medium of education, at least partially, in cognitively demanding decontextualized subjects, at least in grades 8-12.

For MINORITY CHILDREN the mother tongue must function as the medium of education in all subjects initially. At least some subjects must be taught through L1 all the way, up to grade 12, but the choice of subjects may vary. It seems that the following development functions well:

- transfer from the known to the unknown;

- transfer from teaching of a language (as a subject) to teaching through the medium of that language;

- transfer from teaching through the medium of L2 in cognitively less demanding, context-embedded subjects, to teaching through the medium of L2 in cognitively demanding decontextualized subjects.

The progression used for all children in the European Union Schools seems close to ideal for minority children. The progression in relation to the (minority) MOTHER TONGUE is as follows:

1. All subjects are taught through the medium of the mother tongue during the first 2 years.

2. All cognitively demanding decontextualized core subjects are taught through the medium of the mother tongue during the first 7 years.

3. There is less teaching through the medium of the mother tongue in grades 8-10, and again more teaching through the medium of the mother tongue in grades 11-12, especially in the most demanding subjects, in order to ensure that the students have understood, can express and critically evaluate them thoroughly.

4. The mother tongue is taught as a subject throughout schooling, from 1-12.

The progression in relation to the SECOND LANGUAGE is as follows:

1. The second language is taught as a subject throughout schooling, from 1-12.

2. The second language becomes a medium of education already in grade 3, but only in cognitively less demanding context-embedded subjects. Teaching can take place in mixed groups, but ideally together with other children for whom the language is also an L2.

3. Teaching in cognitively demanding decontextualized subjects only starts through the medium of L2 when the children have been taught that language as a subject for 7 years (grades 1-7) and have been taught through the medium of that language in cognitively less demanding context-embedded subjects for 5 years (grades 3-7). Children should not be taught demanding decontextualized subjects through L2 together with children for whom the language of instruction is their L1, before grade 8. In European Union Schools this is mostly not done even in grades 9-12 in compulsory subjects, only in elective courses. (Source: Skutnabb-Kangas 2000).