must be borne in mind, however, that the differences observed depending
on the main teaching language are diluted when cross-referenced
with the two other variables taken into account in this section.
So, the differentiated usages of Catalan and Spanish represented
in graph 11 do not so much reflect the teaching language of the
schools as the type of school in which each linguistic model predominates:
in centres predominantly teaching in Catalan there is a high percentage
of Catalan-speaking pupils, and these are schools in areas where
a high level of knowledge of the language is declared; the centres
that use Catalan and Spanish in more or less balanced proportions
have more Spanish-speaking pupils and are distributed in areas where
Catalan is less well known.
Leaving aside what can be gleaned from a first
glance at the graphs, which is also quite predictable (that is,
the fact that the use of Catalan increases as the knowledge and
presence of this language in the environment increases), the most
important conclusion that can be drawn from the above graphs is
the inequality of conditions for the two languages: the use of Catalan
in contexts that favour it is lower than the use of Spanish in conditions
that are favourable to that language. Note that at centres with
more than 70% of Spanish-speaking pupils, almost 94% of speaking
turns are in Spanish, while at schools where there are more than
70% Catalan-speakers, only 82% of turns are in Catalan.
3.4. The weight
of the variables: mutivariant Analysis
Once the influence of each of the variables on
the linguistic choices of Catalan schoolchildren has been analysed,
it only remains to determine the variables out of the whole set
which are most important in configuring their linguistic practices.
To do this, two explanatory multivariant analysis techniques have
been applied: discriminant analysis and logistical (see a summary
in Galindo 2006).
The application of the multivariant analysis techniques
makes it possible to differentiate two sets of variables: (1) those
taking into account the language of relations with the non-family
and non-formal-education environment and (2) those related to the
family language and formal education. There is yet another group
of variables, (3) those referring to linguistic competence in Catalan
and Spanish, which have not been included in either of the previous
sets because it has been considered that they cut across all the
variables: achieving good linguistic competence does not depend
only on school education but instead is also related to actual use
of the language.
The first set of variables (language of relations
with the non-family and non-formal-education environment) include
the social networks of the speaker and the listener, the linguistic
conditions at the centre and the environmental linguistic conditions.
These variables are the ones that are shown to be more important
in configuring linguistic usages.
The first two variables mentioned (the social networks of the speaker
and the listener) include information declared by the interlocutors
themselves on the language they normally use outside the home; that
is, they are an indicator of perception of own linguistic usage.
That in almost all of the three multivariant analyses carried out
they should constitute the main factors making it possible to explain
the variation in linguistic practices, demonstrates that real and
perceived usages are not going in opposite directions; however,
the language of social networks does not in itself allow an explanation
of all the variation in usages and this indicates that there are
certain mismatches between what people believe they do and what
they actually do, as can be seen in section 2.3.
The third variable, the linguistic conditions at the school, measures
the percentage of Catalan-speakers at the school. We assume that
the more Catalan-speakers there are in the same area, the more use
they will make of Catalan and, therefore, the percentage representing
the linguistic conditions at the centre also measures the opportunities
interlocutors will have to maintain social relations in Catalan.
The fourth variable, the environmental linguistic conditions, refers
to the knowledge of Catalan the residents of the area declare that
they have. As is to be expected, higher percentages of people who
say they understand and know how to speak Catalan in the area correspond
to greater use of the language at break time.
The second set of variables is that related to family and formal
education linguistic usage and includes the family language of the
speaker and listener and the linguistic teaching conditions. The
inclusion of these variables in the discriminant and logistic regression
functions confirms that the home and classroom language plays a
role in configuring linguistic usages, but the position occupied
by these variables in these functions (always below the first group
of factors) indicates that the role they adopt is a secondary one.
It is clear that, in the age sector analysed, the influence of
the home is still important, but it must be borne in mind that children
often spend many hours outside the family nucleus. Meanwhile, the
adoption of Catalan as the vehicular language for teaching in a
good proportion of schools must have affected children's linguistic
practices in some way, but it has not managed to entirely modify
their linguistic behaviour patterns: Spanish has a considerable
presence in school playgrounds and the convergence to Spanish rule
is still partially followed.
As for the third group of variables, related to children's linguistic
competence, the functions deriving from the multivariant analysis
techniques have only included competence in Catalan. This indicates
that competence in Catalan is a factor that configures linguistic
usages: the higher the degree of comprehension in Catalan, the greater
the use of Catalan recorded. This association allows two interpretations:
either the habitual use of Catalan encourages the increase of linguistic
competence in Catalan or only those who have a good knowledge of
Catalan speak the language habitually. By contrast, competence in
Spanish does not allow an explanation of children's linguistic practices:
all of them obtain average or high scores; the fact that they use
Catalan or Spanish must therefore be explained by other factors.
4. Summary and conclusions
The model of joint teaching in Catalan predicted that the extension
of Catalan as the vehicular language for teaching would increase
knowledge and usage of the language among Catalan schoolchildren.
It can now be stated, with few reservations, that the population
of Catalonia who have passed through the education system, at least
in the last 15 years, understands Catalan and can speak it, write
it and read it.