NOVES SL. WINTER 2009. FROM MONOLINGUAL FAMILY BACKGROUND TO BILINGUAL IDENTIFICATION... (2)

While the "attitude" construct has been extensively studied in Catalan socio-linguistics, language confidence has not been dealt with in socio-linguistic studies of Catalan. Confidence is defined as "a relative lack of anxiety when using an L2 coupled with the belief in being able to cope linguistically with the L2 situation" (Rubenfeld et al. 2006). According to various studies by Clément and associates, confidence is a key construct in explaining the use of a second language and the identification of an individual with the second language group. Broadly speaking, Clément's studies have shown the following: "contact with or confidence in a second language leads individuals to identify with the L2 community. This process brings about more positive representations of the L2 culture. On a day to day basis, these studies suggest that learning an L2 may have a positive influence on inter-group relations. In the context of learning an L2, we find a greater identification with that community, which consequently makes us feel more positive about the community" (Rubenfeld, et al., 2006, p. 627).

Lastly, this paper includes the use of language in social networks as a variable which could explain identification with specific languages. As Tabouret-Keller (1997) explains, the psychological component stems from both the relationship with one's parents and the way in which one imitates the behaviour of others. The analysis of social networks provides a relational view of social reality (Molina, 2005), i.e. the relations which individuals have with other individuals around them. While the classical approach to social sciences is based on the study of individual attributes such as gender, age or level of education, the analysis of social networks focuses on relations between individuals. The main subject of study when considering social networks is, then, the nature of the interaction between individuals. These include the existence of a relationship, its strength and its characteristics. Nevertheless, the attributive and relational views are complementary in sociological or psychological analysis.

When we turn to the relationship between language and identity on an institutional level (explicit or implicit legal measures), we need to take into consideration the importance of language policy, both in general and in education, in the socio-linguistic context of Catalonia and the Aragonese border region. The most obvious difference between Catalonia and Aragon is that in the former Catalan is the language of instruction in primary education, while in Aragon it has only a token presence. Other differences are shown in Table 1.

Table 1. Socio-linguistic description of Mataró and the Aragonese border area

	Catalunya / Mataró	Aragonese border area
Population	7.210.508 milions / 119.858	47.686
Political status	Autonomous community	Part of the Autonomous Community of Aragon, with no political status of its own
Legal situation of Catalan	Laws on language policy (1983, 1998)	Not an official language. Recognised in the Law on the Aragonese cultural heritage (1999)
Language policy in education	Language immersion: Catalan is the language of instruction in primary and secondary education, the aim being competence in both Catalan and Spanish.	Pupils are entitled to two hours a week of Catalan as an optional subject at school

2.2. Language and identity in Catalonia

The relationship between language and identity in socio-linguistics in areas where Catalan is spoken is a topic which has been studied extensively from different theoretical points of view, although quantitative studies predominate (see Baldaquí, 2004; Bastardas, 2007; Fabà, 2005a, 2005b; Fundació Congrés de Cultura Catalana, 2006; Llobera, 2001; Pujolar, 1997; Querol, 2001; and Strubell, 2008, among others). The most recent data for Catalonia and the Aragonese border area is that provided by the corresponding statistical studies on language use (EULC 03, Torres, 2005; EULF 03, Sorolla, 2005; EUL 03, Querol et al., 2007). For example, EULC 03 figures for Catalonia show that the majority first language in Catalonia is Spanish (53.5% of the population, as against 40.4% for Catalan), while the majority language of identity is Catalan (48.8% of the population, as against 44.3% for Spanish) (Fabà, 2005a).

Most studies tend to examine two variables to see how the language of identity relates to another socio-linguistic variable. For example, Vila (2006) studies the relationship between language of identity and language use and reports three facts: a) Catalan has a certain power of attraction (a considerable number of speakers have Spanish as their first language but identify with Catalan), b) figures for the use of Catalan are explained more readily by the proportion of speakers who identify with Catalan than the proportion of speakers who have Catalan as their first language and c) language transmission between generations in Catalonia favours Catalan (see also Fabà 2005a, 2005b and Torres 2005).

All the studies carried out up to now deal with samples of young individuals and adults (aged 15 and over in the case of EULC 03 and EULF 03). An exception, both because of its methodology and the age range of the sample, is the study by Baldaquí (2004), who researched the perception of ethno-linguistic vitality by young people (n=202, aged 13-14) in l'Alacantí county. The results of the study showed that the ethnolinguistic vitality perceived by respondents depends on both objective vitality (the sociolinguistic environment) and a series of sociolinguistic variables (first language, educational programmes, attitudes to language, and language learning or conservation). Querol (2001) also carried out multi-variable analyses to explain and predict language use in Catalonia by 16-year-old pupils according to three variables: the social representation of the languages, reference groups and social networks. Results for 2000 show that the main variables which predict the use of Catalan are the representation of Spanish, the social network in Catalan, the social network in Spanish, and identity. It should be noted that results for 1993 included the parents' language as one of the main variables but not social networks in Spanish.

A number of sociolinguistic studies have dealt with samples whose ages were similar to those of the subjects of this paper, although they were slightly older. They focus on the study of attitudes to language and its relationship to sociolinguistic variables (see Lasagabaster, 2003; Lasagabaster & Huguet, 2007, for a summary of these studies). The results given by Huguet (2007) for a sample of 309 university students in Girona and Lleida (average age 20) show that there is a significant difference in linguistic attitudes to Catalan according to the first language of the respondents (more favourable attitudes among those whose first language was Catalan), the language background at school (more favourable to Catalan among those educated in Catalan), and the predominant language in the town or city in which they live (more favourable to Catalan the more widely Catalan was used). Other variables, such as gender, social and professional status, and the population of the town or city, did not give significant differences.

Finally, Huguet and Suïls (1998) studied attitudes to language among the pupils in the second year of compulsory secondary education (the eighth year of basic general education at the time of the study, 1995-1996, n=257) in the border area between Catalonia and Aragon (Baix Cinca and Baix Segre). They found that pupils in both areas had positive attitudes to Catalan and that the variable which could explain differences in attitude was the family language background, even though the situations observed were reversed: attitudes to Catalan were positive in the Baix Segre, while they were positive to Spanish in the Baix Cinca; however, less positive attitudes to Spanish were recorded in the Baix Segre and to Catalan in the Baix Cinca, depending on the first language of the respondents (for example, a less favourable attitude to Spanish among Catalan speakers in the Baix Segre).

In short, previous studies have demonstrated the possible interrelation between different sociolinguistic variables when explaining those variables which are considered primary (or independent). Depending on the methodology of the studies, what has been considered primary has been the use, identity or attitude studied. Today, therefore, we have a jigsaw of sociolinguistic studies which provide multiple sets of data about samples varying widely, both in age and in sociolinguistic background. This study deals with an age range which has not been extensively studied (pupils in the sixth year of primary education) and presents a descriptive and discriminant analysis, which attempts to explain the relationship between the respondents' language of identity and the language they speak with their parents. The fact that the data comes from two areas which are markedly different from a sociolinguistic point of view also enables us to investigate the impact which the sociolinguistic environment has on the language of identity and the language used with parents.