Notó saw that there was a close relationship between the degree of awareness of social identity and level of oral and written ability of Andorrans in their native Catalan. He realised, too, that the greater the level knowledge of oral and written Catalan, the better the knowledge of the other languages taught and used in the Principality: French and Spanish. He could establish quite clearly that it was time to forget once and for all the widespread prejudice and fear that studying more than one language at a time could negatively affect children's learning processes.

Thus, among the most specific aims and objectives in Note's work were the need to examine this collective awareness of identity among young Andorran children, as well as gauge levels of oral and written skills, not only in Catalan, but also Spanish and French in Andorra. Hopefullly, in the medium term, this would lead to the the setting up of a large data source to be able to monitor and evaluate the process of Andorranisation among children and young people of the country.

4.2 Methodology

On the operative level, Notó designed the project to have two stages or phases: firstly administering the questionnaire during school year 1980-81 to all school children in Andorra la Vella (the largest town in the Principality) representing some 40% of the total -and then secondly continuing during school year 1981-82 with the remaining pupils in the country. So in this case there was no sampling, and no qualitative interviews, since the objective was to analyse the entire captive school population.

Note's research was based on a battery of different tests and varied texts which he intended to administer to all elementary secondary school pupils. In practice, however, this was not possible, (9) and he was only able to collect results from children attending Andorran grant subsidised schools as well as some of those attending schools on the Spanish model during school year 1980-81.

4.3 Results and conclusions

In the main conclusions, the author states that the school pupils of Andorra have considerable knowledge of certain specific aspects of Andorran identity, of their history, flag... However, he also notes that, paradoxically, they show little sociocentricity, compared with other communities. He relates the lack of such sociocentricity –a common element of the social identity of people in countries with a more defined political framework -to the co-sovereignty under which Andorra has functioned since the 13^th century, and to the low percentage of native-born Andorrans in the population. One only has to remember the vertiginous demographic growth the country has experienced over the last third of the 20th century, with continuing waves of immigration. This explains why the Principality now has 70,000 inhabitants in contrast to the little more than 6,000 that there were until the sixties.

Around 1958 the transformation of the country's commerce had already begun. A commercial structure was emerging that was vulnerable and yet an attractive magnet for finance. In a very few years Andorran society yielded to the processes of change to tertiary economic activity sweeping aside the little that remained of the old Pyrenees society and the traditional way of life analysed by Stancliff. A country on the border between larger countries, Andorra would from now on offer everything, from relaxation, skiing and tourism, to shopping in extensive commercial centres where, significantly, one can find everything from post cards of the Eiffel Tower, to post cards of (highly Spanish) Sevillanas, taking in pictures of blatant Andorran culture on the way. The sociolinguistic situation of such a society is necessarily complex, even if Catalan is the sole official language...

Notó also noted that in the analysis of comprehension of standard or literary written texts, comprehension of texts in Catalan was usually lower than comprehension of texts in Spanish for Andorran children. It is very likely that this lack of knowledge and greater uncertainty in the case of Catalan bears a direct relationship to the scant presence of Catalan in the classrooms at that time, apart from the fact that probably these texts distributed to the students were very different from one language to another.

It should be pointed out, lastly, that this research was not at all comparable to Lixfeld's work, since however close they may be in time and place, they used very different methodology, and the aims and objectives were different too, as were the segments of the population studied and the collection and analysis of data employed .

5. The Sociolinguistic study by Emili Boix and Jaume Farràs on Usos, coneixements i ideologies lingüistics dels joves de secundària andorrans (Use, knowledge and linguistic ideologies of Andorran secondary school pupils)

5.1 Introduction, justification and objectives

This work was designed and formalised in 1991, under the auspices of the Institut d'Estudis Andorrans, and was based in part on the fascination exercised by Andorra's genuinely bipolar structure -astride the French and Spanish states both geographically and politically. It was also in part a response to the recent drafting of the Andorran constitution, to the existence of large groups of immigrants (constituting the majority of the population) and the consequences that these socio-political and demographic facts could have on the current de facto social multilingualism to be found in the Valleys.

We found it useful to take a close look at Lixfeld and Notó's work on the most notable aspects of the sociolinguistic situation in this tiny country, and the processes of construction of its social, economic, cultural and political identity. In addition to having this information on the main trends in terms of language knowledge and use among the overall population of Andorra, we are also aware of other contributions to our knowledge of the construction of (Andorran) identity among school children. We were interested to know if there was a still a comparative lack of knowledge of (written) Catalan compared with the solid knowledge of Spanish and French exhibited by many inhabitants and students of Andorra. It would be useful, we thought, to observe the phenomena of language contact and interference together with code-switching, common as these are in the day-to-day language use of the streets and squares, the classrooms and families, deriving, as they do from very extensive multilingualism especially among native Andorrans. We also wanted to determine if there are still unfavourable attitudes to Catalan among the foreign population in Andorra, especially in the case of French speakers, and we wanted to take a close look at whether the new waves of immigration (the Gallicians and the Portuguese) are being assimilated, especially via school system.

We wanted, lastly, to find out whether Catalan was functioning as inter-community lingua franca in a country –the only country, in fact!- where Catalan is the sole official language. And yet where, paradoxically, changing circumstances make it not strictly necessary to refer to such sentiments, facts and phenomena in Catalan. This in a country where there are many Catalans and Andorrans, and these latter are virtually the only de facto multilingual speakers in Andorra who are capable of resolving satisfactorily any situation in one or more languages other than their own. Thus they often relinquished their one and only official language in favour of Spanish. This paradox would seem to have perpetuated itself several years later, with the habitual use of other languages also in contact between generations of young people of foreign origin. The presence of a long-established if shrinking number of French speakers in the country, and the influx of Gallicians and Portuguese, did not alter the fact that they all ended up in the playgrounds and corridors speaking Spanish, a language which is not exactly official or even co-official in the Valleys, nor in many cases the language used at home.

Analysing this extensive range of multilingual usage and observing the situations in which this occurred was a difficult if fascinating task. We placed our trust in the statistic torture of matrices of data, and in the qualitative analysis of interviews, to be able to unravel the tangle of reasons, as far as was possible, for this resorting to language switching with such ease.

5.2 Methodology and methods of data collection

The data for this research derive from the quantitative survey of practically the whole of the school population of Andorra between the ages of 13 and 18 years. Containing more than 200 questions, the questionnaire was answered by almost 1,300 students. The research data also includes the materials generated by around thirty semi-directed interviews, carried out with young people from a range of different backgrounds that make up the linguistic demography of Andorra. Both sources constitute the raw material, the corpus with which we worked to produce the different sections of the research.

5.2.1 The survey

The questionnaires were aimed at all eighth year students attending secondary EGB school courses, both FP (non-university stream) and BUP-COU (academic or university stream) students. We wanted to restrict the universe (population) to school attenders and not resort to any sampling technique, since this was, after all, a relatively small population. We accepted that on the date arranged with head teachers for the conducting of the interviews, some students would be absent, or that some group might be away on a school trip or skiing holiday. In the latter case, an effort was made to arrange another date on which to come and give the questionnaire.

The large number of students at the French Lycée, furthermore, meant that it was debatable whether we could interrupt the smooth functioning of the school to carry out an interview with the school's entire student population. The idea of a random sample similarly meant interfering with the class group for the duration of the interview, and this neither we nor the head teachers thought was indispensable nor advisable. For that reason we opted for administering the questionnaire to 14 class groups proportionally distributed across the various years and teaching levels. Normally the questionnaire was given by a member of the research team in almost all instances

There is one more observation to be made regarding the French lycée. The differences between the Spanish and French educational system meant that students of the same age (13-14) had left primary education and begun to attend the Lycée when, according to the Spanish system they were still attending primary school. It was, therefore, to achieve homogeneity of the population by age, that we also interviewed students of the second cycle of EGB in all other schools.

The age segments of students in question and the completed questionnaires are from school year 1991-92 and from the following schools:

5.2.2 Semi-directed interviews

The qualitative research, carried out by means of semi-directed (semi-structured) interviews, was designed to act like a funnel. Starting with very open questions, the interview become increasingly restricted to and centred on very specific behaviour, attitudes and opinions relating to the sociolinguistic situation closest to the interviewee. The age group in question included many of the young people affected by the problem of obtaining of Andorran nationality under the restrictive Law of 1975.

We looked for possible interviewees using the snowball technique until we had interviewed 15 Catalan speakers, 2 bilingual Spanish/Catalan families, 9 Spanish speakers, 1 Francophone, 2 Portuguese speakers, 1 bilingual from a Spanish/Portuguese family and 1 bilingual from a Spanish/French family.

5.3 Results and conclusions

Some 57% of the students had been born in Andorra. One in four of the rest of the students were from Catalan-Valentian speaking areas of Spain. Those born in the rest of Spain constituted less than 10% of the total, and those born in France constituted little more than 3%. Their parents' generation, on the other hand, were newcomers to Andorra in the majority of cases. Only around 10% of fathers or mothers were native-born Andorrans. One out of three of the parents came from somewhere in the wider Catalan-speaking area including Andorra and around 41% were from Spain. Those from France amounted to 4% overall. This amalgam of origins makes it difficult to find offspring of native-born Andorran parents (only 4.5% of our total) or of mixed couples (12.2%) such that more than 80% of the parents of these students were of foreign origin.