Constitutions, Languages, Definitions and the European Charter for Regional or Minority Languages

Under this general title, very different situations can be found. In all of them the issue in question has to do with the separateness of the language, its relation with other possible varieties or languages or its name. These questions are very attached to the identity issues of groups and peoples and, therefore, very sensitive in many particular contexts. We will try to present all cases separating different categories of problems, but recognising that some of the questions interact with each other. The four sections showed below are just an attempt to organise the cases in smaller units, rather than a radically clear cut up classification.

a) Cases related to the identity or name of different or unique languages in relation to political units

Page 239

A first group of problems has to do with the identification of the languages covered by the Charter. In particular, we can list a number of cases where the existence of one or two separate languages can be put into question, whereas in other cases, some people challenge the different denomination of what could be considered as a common language. The point is that these differences of interpretation are linked, to some extent, to political units such as independent states or subnational autonomous territories. Very often, political boundaries create different linguistic identities or different denominations where most of the linguists find a common system.

This is the case for several linguistic situations in Europe. Among others, we can refer here to the division between Moldavian and Romanian languages, Serbian, Croatian and Bosnian languages, or Catalan and Valencian languages. According to the implementation of the Charter, it seems that the Committee of Experts has accepted the division between Serbian, Croatian and Bosnian, on the one hand, and Catalan and Valencian on the other hand. However, it must be clarified that in the first case, the closeness of the languages has also been recognised⁶. In fact, from a linguistic point of view, this separation can be seriously challenged. In the Catalan-Valencian case, the unity of the language has already been recognised, and the Committee of Experts, although maintaining the formal different denomination, reminds of the recognition of language identity between Valencian and Catalan by the Valencian Academy of the language⁷. As for the Moldavian-Romania case, no report has been published in reference to this language so far and therefore there is no clear conclusion in this respect. However, taking account of the aforementioned cases, the Committee may call the language with different names, without showing any clear opinion on the unity of these languages.

Another possible case of identification between languages is that of Czech and Slovak, although in general they are considered as separate languages. The division of the former Czechoslovakia into two different independent states has probably closed the question, although the closeness between the two idioms is also recognised as a factor to explain certain “special status”⁸. Finally, in thePage 240 case of Galician and Portuguese, a common linguistic background can be found and it is admitted from a linguistic approach. However, the separate political development of independent Portugal helped in standardising a different language. Although in the Autonomous Community of Galicia there are social or political sectors willing to underline the identification of Galician with Portuguese, the separateness between the two seems to be widely admitted nowadays. However, at the same time, the closeness between Galician and Portuguese is also officially mentioned⁹.

b) Cases related to the lack of unity of a set of variants/languages/language system

In a second group of cases we will include those situations where the unity or separation of languages or varieties is not expressly linked to political boundaries or units. There is a number of cases in which several dialects, varieties or idioms are to some extent related through a common language system. In some cases they share a common name, whereas in others, specific circumstances have driven to the adoption of different names. Some debates about unity or diversity of languages or dialects do also exist in different European corners. Finally, in other cases, the question has to do again with the name given to the language in different contexts, but in this case with no clear relation to the name of any political entity.

The most complex set of languages or linguistic varieties in Europe is clearly that of the Roma languages. In fact, within the framework of the Charter, several names and denominations are officially used by different countries to refer to the main bulk of this Indo-arian branch of idioms: Roma, Romani, Romany and Romanese. Sweden refers to Romani Chib. Netherlands include also the reference to Sinti along with Roma when referring to the language of this group¹⁰. Norway makes a difference between Romanese, as the language of the Roma and Romani, as the language of travellers¹¹. In other occasions, such as the report concerning the United Kingdom, the reference is made to Romani languages, in plural¹². In the aforementioned case of Sweden, a vari-Page 241ety of romani languages is referred¹³. It is far from clear that all references included to Roma languages in the different members states are pointing our to the same language. In some cases, it is also recognised that some Roma speak a different kind of language, like the Beas in Hungary, which should be considered as related to the Latin branch of the Indo-European languages, and not to the Iranian branch. The unfavourable conditions of the Roma communities across Europe make it more difficult to identify or standardise the idioms they use. Only three countries include Roma languages among the ones protected under part III of the Charter, whereas up to 9 other countries protect them only through part II. Without doubt, this group would be the most complex case when analysing the unity or diversity of the languages covered by the Charter.

A relatively similar situation is that of the Sami languages. However, in this case, the existence of a common filum among different Sami variants is widely accepted. Thus, three member states protect Sami languages through the Charter and they refer to them as Sami (Norway and Sweden), or Saami language (Finland). In the case of Norway, the first report of the Committee of Experts acknowledges three variants of the Sami language: North, Lule and South, which are referred in the instrument of ratification as one same language¹⁴. Within the second report, a fourth variant, Eastern Sami, is also included¹⁵, renamed in the third report as Skolt Sami¹⁶. Also in the case of Sweden, Sami languages are mentioned in plural¹⁷. Similarly, in Finland the Sami language consists of three different variants: Skolt, North and Inari¹⁸. In any case, it is clear that Sami or Saami is considered as one common language in respect of the Charter in each of the countries concerned.

Another set of related languages is that of the Finnic family in respect of Finnish, Meankieli (also called Tornedal Finnish) and Kven (or Kven Finnish). The first two appear protected by the Charter in Sweden, and KvenPage 242 is mentioned in the case of Norway. The closeness of the three variants is not challenged, but both Kven and Meankieli tend to be considered as different languages today, and not only as variants or dialects of Finnish language. It is recognised that Kven Finnish is an old variant of Finnish language¹⁹. The debate on the separateness of this language from Finnish lasted for the two first reports on Norway²⁰. However, after the third report on this country, Kven has been recognised as a separate language²¹. In the case of Meankieli, it is also admitted that it derives from the same roots as Finnish. Meankieli has been recognised by Swedish authorities as a separate language only very recently²². From the available reports it is not possible to get the perception that the Finnish speakers of Finland or Sweden have about the specificity of these two languages.

Norway offers another example of dichotomy between dialects and languages. Both Bokmal and Nynorsk are recognised as two standard written variants of Norwegian languages, and they share a common official status²³. Although it seems that they are referred as variants and not languages, on the other hand, Nynorsk can be regarded according to the Charter as a less widely used official language. This would push us to consider Nynorsk as a separate language rather than a dialect, at least in respect to implementation of the Charter.

The Frisian linguistic group has evolved in three subgroups: North Frisian, Sater Frisian and (west) Frisian itself²⁴. All three have been separated from one another since centuries and developed differently. The result is that they have evolved in different manners to the point that today they are linguistically speaking quite distinct. Therefore, although there is a clear link among the Frisian languages, they must be considered as different languages when implementing the Charter.

Something similar can be said in the case of Sorbian. Again, a geographic distance or isolation of speakers provoked a different evolution of the language.

Page 243

Thus, from the various dialects of colloquial Sorbian two standard languages developed: Upper and Lower Sorbian²⁵. In this case, the closeness of both variants may be higher, but in legal terms they should also be considered as separate languages.

There are other cases of geographic distance can also be included within this category. In the case of Cyprus, Cypriot Maronite Arabic can be easily related to the Arabic language. In fact, according to the Committee of Experts there is no written standard form of Cypriot Maronite Arabic. However, the long geographical and cultural isolation from the Arab-speaking mainland has led to a unique development and independent linguistic evolution of Cypriot Maronite, as to the point that it can be regarded as a distinct language²⁶. Another case of similarity in the distance seems to be that of Croatian and Burgenland Croatian. As the Committee says, “distinction between Croatian and Burgenland Croatian is very recent. Instrument of ratification is the first official reference to this distinction”²⁷. In fact, the Committee of experts uses the two terms interchangeably in the report. The relation between Crimean Tatar (Ukraine) and Tatar (Finland) cannot be determined through the application of the Charter. The linguistic relation between the two is clear, since the Committee recognises that Tatar is one of the Turkic languages²⁸, as it is Crimean Tatar. However, there is no document referring to the close or vague relation of both Tatar languages. The same can be said in the case of Gagauzian (Ukraine) and Turkish (Romania). Taken from granted that Gagauzian is a language corresponding to the western Turkic languages, in principle it seems that they must be regarded as totally different languages under the European Charter.

Finally, another two cases can be mentioned with some slightly different contexts. Here no factor of geographic distance can be observed. However, a debate on the unity or diversity of a language is present and, to some extent, it is related to the name of the languages concerned. Thus, the Committee of Experts recognises that Yezidi and Kurdish were treated in the same group till independence and that both communities speak the same variety of Kurmanji²⁹. The problem in this case seems to be on the very name of the lan-Page 244guage and, consequently, it may look that the unity of the language is recognised, although they use different scripts (latin-cyrilic)³⁰.

The second polemic case in this respect is that of the Ruthenian and Ukrainian languages. Traditionally, Ruthenians were considered as part of the Ukrainian linguistic group³¹, and for a long time the differentiating factor of this minority, if any, was its religious identity. In some cases, the Committee of Experts has recognised that the Ruthenian language suffers from an insufficient recognition of its specificity vis-a-vis the Ukrainian language³². However, the Ruthenian language was codified in 1995³³ and nowadays up to five member states (Hungary, Croatia, Serbia, Slovakia and Romania) have included Ruthenian and Ukrainian as different languages when ratifying the Charter. This seems to be a founded ground to justify the specificity of the Ruthenian language, although in Ukraine this separation seems to be not easily accepted.

c) Cases related to the consideration of the idiom as a dialect or separate language

The third set of questions related to the identity of languages covered by the Charter refers again to the issue of the specificity of some idioms as separate languages or, on the contrary, as dialects of bigger languages. The most common situation is that of a linguistic variant with a low degree of standardisation and political support. In many of the cases, these languages have suffered form the consideration by authorities or societies as patois languages or uncultivated forms of official languages. Some of those included in this category are formally considered as dialects, whereas in other cases they have been included under the protection of the Charter as separate languages.

The number of languages covered by the Charter that can be allocated within this group is considerable. Thus, in Switzerland, according to the experts’ first report, the Yenish speak a non-codified language that belongs to the Germanic family, a hybrid language resulting from a combination of Romani, Yiddish and German³⁴. Another language related to the same family in Switzerland is Walser, recognised as a variant of German language and there-Page 245fore, not as a separate language³⁵. Following with the same branch of Germanic languages, Limburger is included in relation to the Netherlands. However, as the Committee points out, there seems to be uncertainty as regards the position of this language. Limburger is recognised as a language under the Charter by the Dutch authorities, but is not so far recognised by the Dutch language Union³⁶. This factor, however, does not avoid its consideration as a language protected by the Charter.

Within the same linguistic family, Lower Saxon has been referred in plural, as recognising implicitly some internal diversity³⁷. However, the second report reduces its reference to singular³⁸, being these differences a possible consequence of the historical lack of standardisation of this language. A similar and related case is that of Low German, considered as the ancestral language of the German north³⁹. The experts found that Low German was in practice still treated as a variant of German and not as a language in its own rights⁴⁰, what is a clear option in favour of the separateness of the language, considered as regional language within the German instrument of ratification.

Another possible separate language within the Germanic family is Scanian. However, Swedish authorities defined it as a dialect of Swedish and the committee does not seem to challenge this conclusion⁴¹. Finally, in the case of the United Kingdom, the specificity of Scots and Ulster Scots is often challenged. According to the Charter and its implementations, it seems clear that both languages must be considered as such, although their proximity to English has made it difficult for these languages to receive official recognition as separate languages⁴².

Page 246

Moving from Germanic languages to the Roman family, Spain offers a number of cases that are often regarded as dialects and not separate languages. This would be the case of Asturian or Aragonese, although in both cases there is the clear recognition under the Charter of constituting languages⁴³. On the contrary, the Committee denounces that Asturian Galician is not clearly recognised as a variant of Galician as it should be⁴⁴. Far different, in the case of Aranese, this is closely related to Gascon, and in any case it is considered as a variant of Occitan language⁴⁵. Following with the Roman family, as we already mentioned, in Hungary, Beas is the name of the idiom spoken by some 10% of the Roma community, and it is considered as an archaic version of the Romanian language⁴⁶.

Finally, Karelian, a language of the Finnic group close to Finnish, has been recently recognised by Finnish authorities as a separate language to be covered by the Charter⁴⁷.

d) Cases related to the existence of the language due to non-significant presence (already or nearly extinguished languages)

The fourth set of questions that emerge when determining the languages covered by the Charter has to do with the very existence of the language concerned. There are a number of cases where the extreme weakness of a particular language is used to challenge the possibility of considering it as an existing or different language today. These are normally cases where the number of speakers is very small, including languages whose native speakers ceased to exist in the past.

We can include up to 10 possible languages within this section. Among these, only Prussian/Pruthenian has been clearly excluded form the application of the Charter⁴⁸. Similarly, Yiddish is not considered as a language covered by the Charter in the case of Norway due to the lack of speakers⁴⁹ (it isPage 247 covered, however, in Finland, Netherlands, Romania, Sweden and Switzerland). In the remaining cases, although recognising the low figures of speakers, the Committee of Experts opens the door for monitoring protection through the Charter. For example, the Cornish language ceased been used in the 17^th century. However, after a revival occurred during the last decades, today there may be 100 fluent speakers of this language⁵⁰. A similar situation is that of Manx. Its last first language speaker died in 1974, but in 2001 1689 people with knowledge of Manx language were found⁵¹. Another process of recent revival or a nearly extinct language is that of Aragonese in Spain⁵².

Finally, in other cases the committee underlines the low number of speakers, which makes difficult an effective protection by the Charter. Thus, in Armenia, Assyrian, Kurdish and Greek are spoken by very few people and there is no reliable data on the number of mother tongue speakers of these languages⁵³. In the case of Croatia, according to the committee there is evidence of a traditionally present small community of speakers of a language called Istro-Romanian in Istria⁵⁴. And in Spain, the government has recognised that 4000 people speak Calo, a different idiom from Romany⁵⁵.

The second big group of problems related to the interpretation of the first article of the ECHRML has to do with the legal status of the languages concerned. The issue is now the relationship between the languages spoken in a given country and the legal categories foreseen in the Charter. In this respect, legal categories included at the constitutional level of the different countries are also relevant.

As we mentioned in a previous chapter, the Charter includes up to five categories of languages within its first articles: Official language, Regional or Mi-Page 248nority language, Non-territorial language, Less widely used official language and Languages of immigrants. On the other hand, European constitutional comparative law offers us a wider set of legal categories. The legal categories included in the different countries are the following: Official language, State language, National language, Language of the republic, Linguistic regions, Equal (equalised/parificata) status (with the official language)⁵⁶, Language for inter-cultural(ethnic) communication⁵⁷, and Autochthonous language/principal language (lengua propia)⁵⁸.

As the legal categories present in domestic law are not always totally correspondent to those of the Charter, in some cases it is necessary to open space for interpretation. This need of adaptation has also to consider some of the questions and problems of identifying languages that we have already mentioned in the previous section. The final list of questions can be organised into four different topics:

a) Cases related to the condition of official languages under the Charter

In principle, the Charter excludes official languages, since regional or minority languages are defined as being different from official languages of the state. This has so far been interpreted as referring exclusively to those official languages of the whole state, and not to official languages of certain territories (normally autonomous). Thus, languages such as Catalan, Welsh or Gagauzian can (and must) be included under the protection of the Charter regardless of their territorial official status. This point, however, provokes some space for discussion from different perspectives.

According to this reading, any existing traditional language that is not the official language of the whole territory of the State must be covered by the Charter. However, in the case of Denmark, the Committee of Experts has encouraged Danish authorities to include Faroese and Greenlandic within the field of application of the Charter⁵⁹ because they are not mentioned in the national reports. For the Committee it is surprising not to include these languages within the field of application of the Charter, since this is the kind ofPage 249 languages the Charter war created for⁶⁰. The fact that Faroese and Greenlandic autonomous authorities show little interest in extending the protection of the Charter to their languages should not be relevant, once Denmark has become a member state.

A different case is that of Cyprus, having constitutionally speaking two official languages for the whole country. According to the (Greek) Cyprus authorities, Turkish does not have to be included in the application of the Charter because is an state official language. The Committee has however reminded that it must be guaranteed in practice that Turkish is not treated in a less favourable manner than an ordinary regional or minority language in Cyprus⁶¹. In fact, the condition of official language would not exclude the possible consideration of Turkish as a less widely used official language, but in this case the option of including it must be adopted by the state, since it is not based on objective criteria as it happens with regional or minority languages.

The status of less widely used official language is applied to Swedish in the case of Finland, and to Nynorsk in the case or Norway⁶². This is also the case of Romansh and Italian in Switzerland⁶³, although it is admitted that in certain cantons, French and German can also be seen as less widely used languages⁶⁴. The issue of including official majority languages as less widely used languages to be protected by the Charter in some specific regions must be carefully dealt with. In this respect, the difference between constitutional models a), b) and c), explained in section 2 of this paper, should be taken into consideration. The concept of less widely official language seems to be appropriate for situations involved in models a) and b), but it can be more problematic in countries following model c).

b) Cases related to territorial aspects of traditional minority languages

Following with issues related to territorial aspects, a different case can be increasingly observed in the implementation of the Charter. In fact it is taken for granted that some languages are non territorial languages according to thePage 250 Charter, being the Roma family languages the mostly often candidates to this consideration. However, the settlement of many Roma communities across Europe is notorious and more permanent in temporary terms than it was considered in a first stage. The progressive settlement of Roma communities, Jewish groups or even travellers and their languages in given territories puts in question the consideration of these languages as non-territorial in respect to the development of the Charter in the future.

c) Cases related to the condition of non official de facto majority languages

One of the most complicated issues of the application of the Charter in particular domestic situations is that of the condition of non official languages which happen to be de facto majority languages. This situation was probably not foreseen by the drafters of the Charter in 1992, but it appears in several countries, notably former members of the extinct Soviet Union. In fact, Russian language is widely spoken in these countries, whereas it has the condition of official language only in the cases of Russia, Belarus and in the autonomous regions of Crimea, Trandsnistria and Gagauzia. In many cases Russian is considered the language of inter-cultural communication between groups using different languages⁶⁵. The issue becomes problematic in some specific situations. The two former Soviet countries that have ratified the Charter, Ukraine and Armenia, have included Russian as one of the minority languages to be protected by the Charter. In the case of Moldova, one of the big issues looking to ratification of the Charter is that of the status of Russian language. The Charter does not recognise more categories between official and regional or minority languages. In Moldova, Russian is de facto a majority language, apart from the language for inter-ethnic communication, but it is not official language of the state. On the other hand, the national official language is not known by a significant sector of the population, who is Russian speaker⁶⁶. Whereas in Armenia the size of linguistic minorities is significantly lower, in Moldova the amount of citizens not able to use the official language can be as high as one third of the total. This creates a delicate situation, since the Charter obliges to protect all those languages which are not state official, like Russian in Moldova (or other republics). The protection by the Charter of an official minority language is possible through the concept of less widely usedPage 251 official language for some territories. The problem is the condition of Russian language because on the one hand, its recognition as a minority language along with other minor languages can be regarded by some Russian or Ukrainian speakers as offensive and, on the other hand, it obliges to protect actively a language whose status is de facto better than that of the national language. Obviously, the Charter was created to protect and develop the traditional linguistic heritage of the European continent, and not as a tool of political confrontation between diverging national projects.

In any case, further clarification about the position of majority non official languages is required, as well as how to protect weak languages even in the case that they are legally recognised as official.

d) Cases related to the traditional or new condition of the languages

The last relevant issue in respect of the categorization of languages according to the Charter refers to its traditional character. As it is known, the Charter wishes to protect and promote regional or minority languages traditionally used in Europe and it explicitly excludes languages of immigrants from its scope of application.

The question about the “traditionality” of particular languages in concrete circumstances may arise in several countries. Thus, for example, the Committee of Experts has encouraged the Slovenian authorities to reconsider their position on Croatian language on this point and to clarify the issue of the traditional presence of Serbian and Bosnian⁶⁷, and German⁶⁸. Another case under debate is that of the Roma language in Denmark. According to the Committee, the question whether Roma is traditional or not in this country depends on the language spoken for an small Roma community present in Denmark at the end of the 1960s⁶⁹.

The problem in this point is that no clear or objective criteria can be easily admitted. Many processes of immigration will contribute to reinforce the presence of a particular language in a given country and migration flows do notPage 252 occur in a particular moment. They cover normally a wide range of years and it is difficult to determine if a particular language was already present or not in a given territory and when. The exclusion of languages of immigrants from the Charter’s scope of application has justified a kind of a fixed picture of the linguistic situation of each country, but this landscape is to evolve in the near future. In the case of Liechtenstein, it has been admitted that there is no regional or minority language in the sense of the Charter⁷⁰, without consideration to the languages spoken by foreigners or non-citizens. However, in other cases, immigration and a traditional language are directly related, like in the case of Russian language in Finland, where the Committee has decided to consider this language without the distinction of whether its users are new or old⁷¹.

This means that in the case of newcomers adding new speakers to a minority traditional group, they can benefit from the protection of the Charter to that language. Even more, their presence will reinforce the position of that language to be protected under the Charter. Cases like this could occur with Italian speakers migrating to Switzerland, Russian speakers in Finland, Armenia or Ukraine, Serbian speakers in Slovenia or Arabic speaking immigrants in the Spanish enclaves of Ceuta an Melilla. Being this the case, it could seem unfair to exclude from the protection of the Charter the languages of other groups whose presence in the territory of the state is getting longer. It is clear that today’s migratory processes are affecting the traditional linguistic dynamics of many societies, and the idea of European linguistic heritage must also be adapted to this evolution, mainly if these immigrants become citizens of their respective host states.

In any case, the continuous (and growing) presence of former immigrant languages, which have become in part languages of national citizens, will push us to read again the traditional exclusion of what the Charter calls languages of immigrants. The increasing linguistic diversity also makes necessary to reflect on new models of regulating the use of languages in the public domain, and to challenge some traditional concepts such as the ones included in the first article of the Charter.

Page 253