Given the overwhelming focus on processes of second language acquisition by immigrant minority (IM henceforth) groups, there is much less evidence on the status and use of IM languages across Europe as a result of processes of immigration and minorisation. Obviously, typological differences between IM languages across EU member states do exist, e.g. in terms of the status of IM languages as EU or non-EU languages, or as languages of formerly colonialized source countries. Taken from the latter perspective, e.g. Indian languages are prominent in Great Britain, Arabic languages in France, Congolese languages in Belgium, and Surinamese languages in the Netherlands. Most studies on IM languages in Europe have focussed on a spectrum of IM languages at the level of one particular nation-state (e.g. Alladina & Edwards 1991, LMP 1985, Extra & De Ruiter 2001, Extra et al. 2002, Extra & Verhoeven 1993a, Caubet et al. 2002) or on one particular IM language at the national or European level (e.g. Obdeijn & De Ruiter 1998 and Tilmatine 1997 on Arabic in Europe or Jörgensen 2003 on Turkish in Europe).
Few studies have taken both a crossnational and a crosslinguistic perspective on the status and use of IM languages in Europe (Jaspaert & Kroon 1991, Extra & Verhoeven 1993b, 1998, Fase et al. 1995, Ammerlaan et al. 2001). Here we present the rationale, methodology, and outcomes of the Multilingual Cities Project, carried out as a multiple case study in six major multicultural cities in different EU member-states. Our focus will be on the two major private and public domains in which language transmission occurs, i.e. the home and the school, respectively. For a full report we refer to Extra & Yağmur (2004).
The project was carried out under the auspices of the European Cultural Foundation, established in Amsterdam, and it was coordinated by a research team at Babylon, Centre for Studies of the Multicultural Society, at TilburgUniversity in the Netherlands, in cooperation with universities and educational authorities in all participating cities. The aims of the MCP were to gather, analyse, and compare multiple data on the status of IM languages at home and at school. In the participating cities, ranging from Northern to Southern Europe, Germanic and/or Romance languages have a dominant status in public life. Figure 1 gives an outline of the project.
Figure 1. Outline of the Multilingual Cities Project (MCP)
The rationale for collecting, analysing and comparing multiple home language data on multicultural school populations derived from three different perspectives:
Our questionnaire for data collection was designed after an ample study and evaluation of language-related questions in nation-wide or large-scale population research in a variety of countries with a long-term history of migration and minorisation processes (see Broeder & Extra 1998), and also derived from extensive empirical experiences gained in carrying out municipal home language surveys amongst pupils both in primary and secondary schools in the Netherlands (see Extra et al. 2001, 2002). Table 1 gives an overview of the resulting data base derived from the reports of primary school children in an age range of 4-12 years (only in The Hague were also data collected at secondary schools). The total crossnational sample consists of more than 160,000 pupils.
Table 1. Overview of the MCP data base (* Dutch-medium schools only;
On the basis of the home language profiles of all major language groups, a crosslinguistic and pseudolongitudinal comparison was made of the reported multiple dimensions of language proficiency, language choice, language dominance, and language preference. For comparative analyses, these four dimensions have been operationalized as follows:
The operationalization of the first and second dimensions (language proficiency and language choice) was aimed at a maximal scope for tracing language vitality. Language understanding is generally the least demanding of the four language skills involved, and the mother acts generally as the major gatekeeper for intergenerational language transmission (Clyne 2003). The final aim was the construction of a language vitality index (henceforward LVI), based on the outcomes of the four dimensions presented above. These four dimensions are compared as proportional scores in terms of the mean proportion of pupils per language group that indicated a positive response to the relevant questions. The LVI is, in turn, the mean value of these four proportional scores. This LVI is by definition an arbitrary index, in the sense that the chosen dimensions with the chosen operationalizations are equally weighted.
The outcomes of the local surveys were aggregated in one crossnational home language survey (HLS) data base. Two criteria were used to select 20 languages for crossnational analyses: each language should be represented by at least 3 cities, and each city should be represented in the crossnational HLS data base by at least 30 pupils in the age range of 6-11 years. Our focus on this age range was motivated by comparability considerations: this range is represented in the local HLS data bases of all participating cities. Romani/Sinte was included in the crossnational analyses because of its special status in our list of 20 languages as a language without territorium status. Two languages have an exceptional status: English ‘invaded’ the local HLS’s as a language of international prestige, and Romani/Sinte is solidly represented in Hamburg and Göteborg only.
In the crossnational and crosslinguistic analyses, three age groups and three generations are distinguished. The age groups consist of children aged 6/7, 8/9, and 10/11 years old. The three generations have been operationalized as follows: G1: pupil + father + mother born abroad; G2: pupil born in the country of residence, father and/or mother born abroad; G3: pupil + father + mother born in the country of residence. On the basis of this categorization, intergenerational shift can be globally estimated. In Table 2 we present the language vitality indices (LVI) of the combined age groups (6-11 years) per language group in decreasing order.
Table 2. LVI o. f combined age groups (6-11 years) per language group in decreasing order (derived from Extra & Yağmur 2004: 375)
Romani/Sinte was found to have the highest language vitality across age groups, and English and German had the lowest. The bottom position of English was explained by the fact that this language has a higher status as lingua franca than as language at home. The top position of Romani/Sinte was also observed in earlier and similar research amongst children in the Netherlands, and confirmed by various other studies of this particular language community. One reason why language vitality is a core value for the Roma across Europe is the absence of source country references as alternative markers of identity –in contrast to almost all other language groups under consideration.
There are strong differences between language groups in the distribution of pupils across different generations. In most language groups, second-generation pupils are best represented and third-generation pupils least. In conformity with expectations, the obtained data finally show a stronger decrease of language vitality across generations than across age groups. The strongest intergenerational shift between first- and third-generation pupils emerges for Polish, whereas the strongest intergenerational maintenance of language vitality occurs for Romani/Sinte and Turkish.
The local language surveys have delivered a wealth of hidden evidence on the distribution and vitality of IM languages at home across European cities and nation-states. Apart from Madrid, latecomer amongst our focal cities in respect of immigration, the proportion of primary school children in whose homes other languages were used next to or instead of the mainstream language ranged per city between one third and more than a half. The total number of traced other languages ranged per city between 50 and 90; the common pattern was that few languages were often referred to by the children and that many languages were referred to only a few times. The findings show that making use of more than one language is a way of life for an increasing number of children across Europe. Mainstream and non-mainstream languages should not be conceived of in terms of competition. Rather, the data show that these languages are used as alternatives, dependent on such factors as type of context or interlocutor. The data make also clear that the use of other languages at home does not occur at the cost of competence in the mainstream language. Many children who addressed their parents in another language reported to be dominant in the mainstream language.
Amongst the major 20 languages in the participating cities, 10 languages are of European origin and 10 languages stem from abroad. These findings show that the traditional concept of language diversity in Europe should be reconsidered and extended. The outcomes of the local language surveys also demonstrate the high status of English amongst primary school children across Europe. Its intrusion in the children’s homes is apparent from the position of English in the top-5 of non-national languages referred to by the children in all participating cities. This outcome cannot be explained as an effect of migration and minorization only. The children’s reference to English also derives from the status of English as the international language of power and prestige. English has become the dominant lingua franca for cross-national communication across Europe. Moreover, children have access to English through a variety of media, and English is commonly taught in particular grades at primary schools.