DEMOGRAPHY VS. LEGITIMACY: CURRENT ISSUES IN SWISS LANGUAGE POLICY

This paper discusses current language policy debates on national and immigrated languages in Switzerland. Problems with the principle of territoriality, which represents a locally monolingual regime in an officially quadrilingual country, and other issues related to the legal status of languages are discussed. The proportional representation of the national minorities and the use of their languages in particular contexts such as the federal administration or the army is discussed, as well as the current debate on which foreign languages should be given priority in compulsory primary education. Drawing on language ideology research, the contribution shows how specific aspects of linguistic diversity are focused and addressed in particular contexts (e.g. national languages, standard languages), whereas others are backgrounded, denied legitimacy or simply erased (e.g. immigrated languages, dialects). The discussion addresses also the demographic weight of the languages and varieties in Switzerland as well as in the world and uses census data to illustrate the stability and changes regarding the official and immigrated languages across time.


PRELIMINARY REMARKS 1
Linguists are not immune to language ideologies. Linguists' "language regard" (Preston 2010), i.e. the way we mentally construe languages and varieties, is an important part of the public discourse on language matters. Funding in applied linguistics and language teaching also depends on the relevance attributed to questions of language management by those who provide the funds. We consider our topic socially pertinent and generally do not complain about it being a subject of political and institutional debates. Benefitting from public funds for research on the status and the teaching and learning of national and non-national languages, however, does not absolve us from a minimal scholarly research requirement: awareness of the ideological underpinnings that interfere with our research. This is particularly important since language policy discussions often involve moral stances, as indicated by the term linguistic human rights advocated by some scholars (Skutnabb-Kangas, diversity within 'languages'. Considerable sums of money are spent on the documentation of dialects both on the Romance and on the Germanic side. 2 On the other hand, some consider the vitality of the Swiss German dialects a threat to national cohesion, and they even consider the use of Alemannic dialects by politicians in public illegal (Ribeaud 2010). The European Charter for the protection of regional and minority languages (Council of Europe 1992), which Switzerland ratified in 1997, explicitly excludes dialects from its protection. Current debates on adding the Franco-Provençal patois, still spoken as a first language in some parts of Western Switzerland (Elmiger et al. 2012;Kienzle 2011), to the list of minority languages, raise the question whether these varieties should be granted language status. In the canton of Valais, at least in the view of some prominent actors (Bernard Bornet, 3 personal communication), the status planning efforts for the patois are explicitly tied to an equal protection of 'Oberwallisertiitsch', the Alemannic variety spoken in the upper part of the valley.
The Romansh situation is at least as complex. Traditionally, philologists distinguish five different 'idioms' (≈varieties) of Romansh, all of which have a more or less standardized spelling and written tradition (Liver 1999). There have been several attempts at a common standard language, the latest being the introduction of Rumansch Grischun (RG) in the administration, media and, with rather variable success, in school (Coray 2008;Darms 1993;Berthele 2015b). Whereas the use of RG in the federal and cantonal administration is uncontroversial, the use of this common standard in school, in particular as the language of pedagogical material and in oral instruction, is extremely controversial and the last attempt to impose it failed in most areas (Berthele 2015b;Lindt-Bangerter and Berthele 2009). As we will see below, the demographic situation of Romansh is such that all speakers of Romansh today develop a high level of proficiency in Swiss and Standard German.

TERRITORIALITY AND THE LANGUAGE CURRICULUM IN SWISS PRIMARY SCHOOLS
Linguistic status planning in Switzerland is based on the principle of territoriality. This principle is constitutionally enshrined, for example in article 70 of the constitution, which stipulates that the cantons should respect the "traditional territorial distribution of languages". The principle of territoriality, to put it bluntly, works well where status planning is unnecessary, i.e. in areas where there is only one dominant traditional language. In border zones, e.g. in the bilingual cantons of Fribourg or Valais, a strict application of territoriality erases the existence of traditional minorities who happen to live on the 'wrong' side of the language border. Accordingly, the Swiss supreme court has repeatedly confirmed the right of children to be schooled in the locally weak language in towns close to the border (e.g., Altermatt 2003; Angst and Rietiker 2015: 124). In another region, the principle of territoriality is used in a rather unorthodox manner: The language law in the trilingual canton of Graubünden imposes Romansh or Italian as the sole official languages in administration and school in places where at least 40% of the population uses the minority language (Art. 16 of the law from the year 2006).
A typical feature of Swiss status planning is the periodically surging debate on foreign languages in the obligatory school curriculum. Because Switzerland is a federal state, education and culture are in the competence of the cantons. Nevertheless, there are attempts at 'harmonizing' learning outcomes and some main cornerstones of the curricula in the country, licensed by a constitutional article (62.4) voted by the Swiss people in 2006. One aspect of this harmonization entails the introduction of two foreign languages in primary school, one of which should obligatorily be a national language.
This endeavor meets resistance for several reasons. First, some people believe that introducing two foreign languages in primary school overemphasizes languages at the expense of other subjects. Thus, they advocate introducing the second foreign language in secondary school only. A recent example here is the referendum launched in 2013 and declared void in 2015 in the trilingual canton of Graubünden. This referendum, similarly to referenda in other cantons, aimed to limit to one the number of foreign languages taught at primary level. 4 Second, some people fear that two foreign languages is too heavy a burden for weak students, and would therefore like to reduce the minimal language requirements. Third, some oppose the system's flexibility and want to guarantee that the first foreign language taught is a national language -the rationale being that the national languages need to be protected against the dominance of English. Lively debates are currently going on, with popular votes in several cantons on these matters. This controversy replicates a debate that took place around the year 2000, when English was the first foreign language taught in a reform pilot in the canton of Zürich (Mittler 1998).
Further arena in which the status of the national languages is hotly debated are the federal administration and the Swiss army. Legal dispositions (above all the revised language 4 The German-speaking children of the canton would have been taught English, while the Romansh and Italian-speaking children would have been taught German as a foreign language. This initiative was declared null and void by the local parliament on the grounds of its incompatibility with constitutional provisions regarding the equality of educational opportunities and regarding the permeability within the Swiss educational system (moving from one territory into the other would entail completely different language curricula; see section 4 for further discussion).
Ordinance from 2014) impose a proportional representation of the national languages in the federal administration (68.5-70.5% German; 21.5-23.5% French; 6.5-8.5% Italian; and 0.5-1% Romansh): the aim is to mirror the demographic weight of the country's different linguistic communities in the Swiss federal administration. Given the differences in size of the communities, it is obvious that barring strong counter-action, the tendency towards the use of German as the working language, as the original language of documents, as well as a tendency to hire German speakers to facilitate communication is unavoidable. Language becomes a selection constraint, and many consider the current policy inefficient when it comes to counteracting the overrepresentation of the German-speaking majority -especially for the top positions in the federal administration.
The resemblance between the debate on the status of German in the Swiss administration and the status of English in the European administration is striking. Several researchers have investigated practices in the administration, from questions relating to mono-or multilingual editing of legal documents (Grüter 2014) to representational aspects and the recruitment mechanisms in several government agencies (Coray et al. 2014;Kübler 2013). The linguistic aspects of current developments in the Swiss army, including the fate of national minorities in increasingly mixed-language branches of service, have also been investigated (Kreis and Lüdi 2009;Berthele and Wittlin 2013;Wittlin 2011). Overall, these investigations show that the traditional methods of language management, most importantly the territorial logic of handing down autonomy to areas and institutional entities that are considered monolingual, are inadequate in times of increasing cooperation among the cantons (coming-together federalism). In other words, as soon as there is a tendency to give more power to the central administration, and to harmonize policies that used to be completely regionalized, a hands-off policy that does not protect the rights of the members of the minority language communities bears the risk of empowering further those who are already powerful (the demographically dominant speech community).

STATUS PLANNING IN SWITZERLAND: SOME PERSONAL OBSERVATIONS
It is impossible to give a full account of the public discourse on national languages and varieties here. One common denominator is that virtually all actors stress their favorable attitude towards multilingualism and respect of languages (and sometimes dialects; cf. Babylonia 3/2014 for an overview of the historical background of the Swiss discourse on the national languages). At the same time, this celebratory discourse of Swiss multilingualism is always selective (Berthele 2014) and disregards the presence of certain languages or dialects which are simply forgotten, considered irrelevant or illegitimate, while the languages or varieties that carry the focus of the particular interest groups are highlighted.
Several typical problems arise in the context of official status planning in Switzerland.
Erasure: Firstly, non-national languages (see section 3 for demographic weight of some of these languages) are often absent from the debates (see Irvine and Gal 2000, on erasure in language ideology research). The minority language policy of the country is strongly focused on national minorities and therefore overlooks the comparatively high proportion of more recently immigrated minorities (see section 3.2 below). Whereas the four national languages are explicitly attributed legal status in the federal legislation, the constitution also protects the "freedom to use any language" (Art. 18). At the same time, no mention of representation of languages other than the national languages is made, e.g. in the context of the representation of the population in the federal administration. When other languages are mentioned, they are often considered a threat to the 'bond' of national languages (i.e., English, Berthele 2001) or as an educational problem, in the case of migrants who do not master the school language (Grossenbacher & Vögeli-Mantovani, 2010;cf. also Esser, 2006 for the German context). There is educational planning for English with respect to foreign language instruction in school, but in other public domains there is no status planning for English in Switzerland. Given the importance of the language in many sectors of public and private life, this is considered potentially problematic by many (Achermann and Künzli 2009).
Proportionality and language use: Second, the idea of proportional representation of the communities in the administration can be challenged. Given the small size of the Romansh and the Italian communities, one could also argue that an overrepresentation is necessary to attain a minimal presence of the minority languages in the administration. Moreover, the mere presence of speakers of a particular minority in the administration does not guarantee that their native languages are actually used.
As argued in Grin (2008) and Berthele and Wittlin (2013), an interesting language regime in multilingual institutions could be receptive multilingualism: Instead of imposing either one common language or translation to and from all languages (the 'panarchic' model), polyglot dialogue (Posner 1991) is practised in some institutional contexts (e.g. the bilingual University of Fribourg's committees and councils). This regime involves people using their respective native or high-proficiency languages in production while having receptive proficiency in two or three other national languages. Receptive multilingualism as a practice could ease the pressure on the school system since developing productive skills such as oral fluidity in several foreign languages requires massive pedagogical investment that the current system seems incapable of providing. Developing receptive skills in several varieties and languages, however, is much cheaper in terms of pedagogical investment and could allow more respect for minorities (Pandolfi, Christopher, and Somenzi 2014;Müller et al. 2009, Gross et al. 2015. School as a battleground: Third, given the reduced importance of traditional institutions emblematically representing national cohesion (e.g., compulsory army service for all Swiss males or a year as an au pair in the francophone part for German-speaking females), school is considered to be the institution that generates national cohesion by teaching national languages. Thus, the educational system that was traditionally regionally anchored and monolingually construed needs to be reinvented (at least according to some vocal participants in the public discourse; for an example see Beacco et al., 2010: 16). It is far from certain whether publicly-funded schools, within the current financial, temporal and personal constraints, can live up to these expectations. In conjunction with the currently rather successful nationalist and anti-European political tendency in Switzerland (in line with tendencies in other parts of Western Europe), language issues regarding compulsory education provide an ideal battleground for political struggles. The rapid changes in the ideological view of dialect in kindergarten and primary school illustrate these struggles: dialect, not long ago, was considered an obstacle to literacy and was banned from the curriculum, whereas, at least in some cantons, it is now compulsory for kindergarten, since it is a feature of national identity (Berthele 2010).

DEMOGRAPHIC WEIGHT OF LANGUAGES
In this section, the emphasis lies on the demographic weight of a selection of languages and varieties spoken in Switzerland. As discussed in section 2, in Switzerland, as in many other countries, the attribution of status to languages is based on the distinction between traditional and immigrated languages. The point in history that is required to become a traditional language is not spelled out clearly, and as Pavlenko (2011) has shown for the Baltic states, it can also depend on particular historical contingencies whether a resident minority and its language is considered legitimate or not.

MAIN LANGUAGES IN SWITZERLAND AND WORLDWIDE
In the case of Switzerland, the status of the traditional languages is controversial in two respects. First of all, the term 'German' is unclear: does it include Swiss German or not? Does the constitutional provision on German as a national language refer to the Swiss variant of Standard German, or to Swiss German dialects, or both? Is Swiss German actually the more traditional, more authentic 'German'? Second, the status of Franco-Provençal, in my view, needs to be clarified: should it be protected by the Charter or not? Is it a dialect of French or a 'real language'? In a first step, I describe the main languages used in Switzerland by charting their demographic weight within the country and in the world (see Figure 1).  Filled squares represent languages that benefit from official support on the federal level, either due to their constitutional status as national languages (German, French, Italian, Romansh) or due to their protection due to the European Charter and the European Council's Framework Convention (Yenish, cf. Hofmann et al. 2015). Empty circles represent languages or varieties that are not explicitly protected by such documents (immigrant languages, but also Swiss German and Franco-Provençal).
Several caveats must be offered: As shown by Brizić (2007), categories such as Turkish in survey or monitoring data are problematic, since a large proportion of emigrants from Turkey either deliberately or inadvertently miscategorize themselves as speakers of Turkish, even though their dominant or native language may be, e.g., a variety of Kurdish or one of many other minority languages spoken in Turkey. The estimates of numbers of speakers provided by the Ethnologue (Gordon 2005) are often also problematic, since the whole enterprise of counting languages and speakers per language is far from being trivial (Mühlhäusler 1996). Furthermore, the term 'main language' in the Swiss census is again rather fuzzy, probably covering notions such as self-assessed dominance, chronology of learning (first language), and local ethnolinguistic vitality of a language in a given context. And finally the figure provides overlapping estimates of both German as a collective term and of Alemannic, the latter also being included in the count of speakers of German.
The figure can be interpreted with respect to several different reference points. One possibility is the comparison of demographic weight on the global (y-axis) and on the local (x-axis) scale. Before we distinguish big and small languages, it is worth pointing out that all languages included in the graph, down to Romansh, can be considered typical to big European languages if we choose the median size of European languages as a reference point (35,600 speakers according to Ethnologue, as opposed to 950 speakers per language in the Pacific; indicated by the dotted line in the detail plot un the upper right corner of Figure 1). There are locally very big languages such as German or its subcategory Alemannic. The locally dominant language German, on the global level, plays more or less in the same league as the locally much less important languages Turkish, Italian, or French -demographically speaking. 6 There are languages that are globally and locally weak in terms of speakers but still have status (Yenish, Romansh). There are locally weak languages that are also globally comparatively small (Albanian, Serbian, and Croatian) and have no status, because they are not territorial languages of Switzerland. At least one language in fact would be traditionally Swiss, but it has only very few speakers -locally and globally -and no status (Franco-Provençal). Finally, the globally strongest languages (Spanish, English, and Portuguese) are locally demographically relatively weak (between about 150,000 and 300,000 speakers), although not as weak as Romansh (around 40,000 speakers), and they have no official status.
The numeric relations depicted in Figure 1 are an emblematic illustration of some of the challenges confronting Swiss linguistic status: as soon as the global level is taken into account, the demographic weight and therefore also the potential usage contexts (cf. the 'Qvalue', de Swaan, 2004) of languages such as English, but also Spanish and Portuguese, is beyond any comparison not only with Romansh, but also outside the range of 'big' European languages such as French or Italian.

REGIONAL DEVELOPMENT OF THE LINGUISTIC DEMOGRAPHY
There are detailed analyses of the demographic situation of the national languages in the different areas of the country (Lüdi and Werlen 2005). The federal statistical office provides detailed data and analyses on its online platform. These data shed light on the current state and on the development of the linguistic situation from a macro level. Figure 3 show the diachronic development of the main languages of the population for the major cities in the Italian-, German-, and French-speaking parts of Switzerland (Lugano, Zürich, Geneva) between 1970 and 2000, based on data from the database STAT-TAB made available by federal statistical office. 7 In Figure 2, the y-axis represents the proportion of the total population that Italian speakers represent, in Figure 3, the y-axis represents the proportion of French speakers.

Figure 2 and
The two figures show the proportion of speakers of Italian and French as well as of other non-territorial languages on the y-axis. Figure 2 shows that the proportion of Italian speakers outside the Italian-speaking territory decreases (line moving downwards for Zürich and Geneva).

Figure 3 shows that the proportion of French speakers in German-and Italian-speaking
Switzerland is stable across time. Within their territories, languages either remain stable, or, as in the case of Geneva and the French-speaking districts overall, French is even strengthened across time ( Figure 3). As Lüdi & Werlen (2005: 29) argue, there is no evidence for massive country-internal migration of German speakers into the Italian or French territories. However, it is clear that the territory in which Romansh is the strong language is gradually shrinking, and actually has been shrinking for centuries now (see Liver 2000). 8 On the other hand, the statistical data show that in the French-and German-speaking territories, the proportion of speakers of languages other than the territorial languages increases, whereas it decreases in Italian-speaking Switzerland. The increase in the presence of other, non-national main languages is as yet not really acknowledged by federal language planning. Obviously, there are recommendations, and there are school-level and regional policies, e.g. with respect to heritage language classes (Giudici and Bühlmann 2014), and the state secretariat for migration (SEM) is active in educational language planning for migrants to develop skills in the national languages, i.e. in an assimilatory perspective. Otherwise no coherent federal language policy taking into account the languages of migration is currently implemented. Meanwhile, as studies on institutional contexts show (e.g. Berthele and Wittlin 2013, for the Swiss Army), English and other languages are used, together with the national languages, if understanding via national languages is difficult.

INDIVIDUAL MULTILINGUALISM
To state the obvious, nobody doubts that official multilingualism does not guarantee comprehensive collective individual bi-/multilingualism. Although macro-level policy can also be an end in itself (e.g. as a political symbol of respect for minorities), the question of the extent to which multilingual policies coincide with the development of individual multilingual repertoires deserves attention. In this section I therefore focus on some evidence on individual multilingual repertoires. Werlen et al. (2009) present data from a representative survey on language proficiency of the adult population in German-, Italian-and French-speaking Switzerland. The results show that the highest proportion of multilingual individuals is found in the Italian-speaking territories, whereas the highest proportion of people who do not speak any language other than the territorial one is found in the Francophone area. There are also some other interesting differences that emerge from theses analyses, e.g. that the highest proportion of speakers who only master English as a foreign language (and not any other national language) are the Francophones (see Werlen 2009, for a brief overview of the results of this survey).
Additional and more recent evidence for individual bi-or multilingualism can be gathered from census data. In Figures 4-6 I again use survey data that can be downloaded from the data base made publicly available by the Swiss federal statistical office, in this case from the 2013 census. The responses stem from the adult population (over 15 years old), and the participants in the survey could indicate more than one main language. This makes it possible to estimate the degree of self-assessed bi-and multilingualism in the Swiss population. As can be seen in Figure 4, the relative proportion of people indicating two or three main languages is higher in the younger population than in the older population. If these differences across age groups indeed are due to a sustained trend towards more individual multilingualism, then the complaint that the Swiss somehow retract into their territorial monolingualism (e.g. in Ribeaud, 2010) cannot be confirmed. The languages in people's repertoires may not be 'only' national languages, and the proficiency in the national languages may not be what it should -according to the expectations of policy makersbut the younger generations display language repertoires certainly not smaller than those of the older generations. When assessing the degree of bi-or multilingualism of a whole population, it is important to distinguish social stratification of the phenomena under investigation. Again, the data depicted in Figure 5 and in Figure 6 only yield limited insights, since they do not allow focus on the individual languages that are part of these repertoires. Nevertheless, Figure 5 shows that there is a slight sectorial and socially stratified tendency towards more bi-and multilingualism in the top segment of the occupational scale (managers), and then again in the domain of sales and other occupations at the lower end of the scale. Corresponding to the stereotype, the agricultural domain correlates with the highest proportion of monolingualism. And, again unsurprisingly, unskilled occupations coincide with relatively high proportions of multilingualism. Figure 6 corroborates this U-shaped relationship in the proportions of biand multilingualism: values are lowest in the group holding intermediate degrees (secondary education on level II), while both lower and higher degrees come with more bilingualism (and, in the case of the tertiary degrees, also multilingualism).
Thus, when making claims about individual linguistic repertoires and educational and occupational success, it seems important to keep in mind that there are at least two forms of bi-and multilingualism. To put it simply, multilingualism within the lower strata of the population typically involves proficiency in languages such as Portuguese, Serbian or Croatian, or Italian, in addition to the local language. As for the elite, the languages typically used are the national languages as well as English (see Lüdi and Werlen 2005: 47). The social connotations of these two types of bi-and multilingualism are different (see, e.g., Berthele, 2012or Imdorf, 2008 for investigations of the effects of ethnicity and bilingualism on the assessment of language proficiency).

CONCLUSIONS -ON THE SELECTIVE CELEBRATION OF LINGUISTIC DIVERSITY
The main goal of this selective overview of Swiss language issues was to confront the traditional, relatively well-established principles, such as territoriality, with the tensions that emerge due to various changes in modern Swiss society. A main characteristic of Swiss (or probably any) public language discourse is the biased focus on particular languages and types of bi-or multilingualism. The celebratory perspective on Swiss quadrilingualism erases not only internal heterogeneity (dialects, patois, etc.), but also the presence of a large number of other, less legitimate languages in the country. This leads to a skewed view of bi-and multilingualism that is prototypically associated with proficiency in legitimate languages. As the census data show, bi-and multilingualism is relatively frequent in socially low strata. However, the celebratory view of multilingualism generally does not take into account that for an important part of the Swiss population, being multilingual can also be regarded as a feature of being underclass, and migration status and lower classes are notoriously confounded in many Western countries. An official policy that generally 'values' multilingualism, but merely focuses on national languages, is unlikely to change the selfperception of these multilinguals. 9 The rather selective attribution of status to languages leads to characteristic tensions between linguistic vitality and language status, two of which I would like to mention in these final considerations: First languages that are not territorial languages: If the first language and its development really have the importance for educational success that is claimed in official documents (see e.g. EDK 2004: 2), then there needs to be an obligatory slot in the curriculum that is dedicated to the first languages of all children, both Swiss speaking non-territorial languages and the immigrant communities (as is practised in some schools in Basel, cf. Luginbühl 2003). Such compulsory and integrated first language instruction needs to be funded and quality monitored by the Swiss school authorities instead of depending largely on 9 The European language portfolio contains elements that are intended to counteract such biases towards particular (legitimate) languages. However, the implementation of portfolio-related activities in Swiss schools are met with merely modest enthusiasm, as surveys have shown (Wokusch 2010, 21).
the embassies of the countries of emigration, as is currently the case. If, however, this importance of the first languages is not as decisive as the official documents state, then it might be better to refrain from mentioning them in language policy documents, in order to avoid paying lip-service only. The current practice offers rather variable access to L1 instruction based on local availability, and there is limited pedagogical control since the classes are funded and organized from a third party. There is a risk that the current half-baked system creates even more inequalities within the educational system.
English: Some scholars regard the lack of status of English as a problem (see section 2.3). The superior status of English as an international language is overtly acknowledged by Swiss language rights specialists (see the quote by Previtali 2013, below). Thus, granting English the status of a working language in administration and education could be an option that would adapt the law to practices currently in place anyway. 10 A hands-off policy regarding the status of English may not cause any serious problems in the near future. However, since it is rather unlikely that the use of English will decrease, such a policy leads to a widening gap between official language status and actual practices. The traditional territorial policy of imposing the use and mastery of the local language for granting residency and citizenship is loosened anyway as soon as the expected tax revenues are sufficiently high, as a recent case in the canton of Zug shows (cf. an article in TagesAnzeiger, 16.4.2014).
Competition between languages is normal and typical for a multilingual state. The rather dramatic differences in size of languages and varieties displayed in Figure 1 mirror the amount of effort that is required to improve the status of languages such as Italian, let alone Romansh: Their communicative value is decreasing in the face of the increasing presence of speakers of global languages such as Spanish, Portuguese or English. The standard response of most vocal advocates of Swiss bi-and multilingualism to this competition is twofold. First, there is an appeal to identity related values: French, Romansh and Italian are 'Swiss' languages, whereas others are not (e.g. Ribeaud 2010). Second, there is a habitual reference to 'explosion theories' of multilingualism, i.e. with theories claiming that learning a third or fourth language is easier and faster than learning a second language (e.g. Hufeisen and Neuner 2004). Thus, according to this view, the debate must not be English or French in the curriculum in German-speaking Switzerland, but English and French. But then Italian has not even entered the picture, let alone Romansh and Spanish. The time that can be allocated to language learning in compulsory school is limited and not all students are language fanatics; nor are all of their teachers. Competition between languages is thus unavoidable, and there is also no doubt that languages, in the view of language users, are regarded as being profoundly unequal (despite the sustained efforts of imposing an egalitarian view on languages, see e.g. Krumm 2003, 39). The example discussed briefly in section 2.2, the referendum declared void by the Graubünden parliament in April 2015, was considered unconstitutional, among other things, because not teaching English as a foreign language to Romansh-and Italianspeaking children means depriving these two minority groups of an important resource: Nel caso dell'adozione dell'iniziativa popolare qui esaminata ci si troverebbe quindi nella curiosa situazione che agli allievi appartenenti alla maggioranza linguistica tedescofona che possiedono quindi già la principale lingua nazionale del paese, sarà anche garantita la possibilità d'apprendere precocemente la lingua internazionale più importante, l'inglese. (Previtali 2013, 19) 11 This clearly shows that, indeed, the languages in the curriculum are considered as being unequal due to differences in national or international status, as we have already illustrated by the sheer differences in numbers of speakers in Figure 1. Along the same lines, Italian as a foreign language as an optional subject in public school in Switzerland, as a second example, is currently not under pressure from French or English (they come first anyway), but from the internationally important language Spanish. To counteract these tendencies via status planning is far from easy. As I have argued in section 2.3 lowering the expectations and focusing on comprehension skills first might be a way to reduce the learning load in the language domain while actually going some way towards increasing the potential usage contexts of minority languages.
This chapter uses some current debates on language management in Switzerland to illustrate the tension that emerges between a national language policy that attributes language rights to speakers of 'traditional' languages, and demographic challenges due to immigration and country-internal migration. Country-internal migration creates a problem mostly for the status of Romansh, due to the emigration of the Romansh speakers from their traditional territories while German speakers immigrate into this territory. In all linguistic regions, immigration of speakers of non-national languages raises the question of the language rights of these speakers. The principle of territoriality is the legal instrument that can be used as a management tool for the protection of traditional languages. However, the legal dispositions 11 In the case of the acceptance of the popular initiative under investigation here we would be in an odd situation in which the pupils who belong to the German-speaking majority, who therefore are already proficient in the main national language, would be granted the opportunity of an early learning onset of the most important international language, English. (translation RB) and language management measures presented and discussed in section 2 are not welladapted to the challenges presented by the demographic changes presented in section 3, neither is the celebratory discourse on multilingualism very helpful for the educational challenges that need to be met. Moreover, high-level language labels such as 'German' or 'Romansh' erase within-category variation that equally gives rise to serious challenges for a consistent diversity-oriented language policy.