These terms mean approximately what they
suggest us at first sight: normal = 'habitual', separation = 'breakage' from
were it was habitual. Warp = 'divergence' and compromise = 'convergence'
with that different from itself. We can say that the warp fact will act when the
speaker has a representation, a perception, more favourable from language B, this
means, from the different language at the normal situation (A): so it produces, a
warp respect what is habitual.
About the compromise
fact, we can say that it will act when the speaker has a different social net to the
first speaking language, and a favourable perception of this other social net. With such a
thing, convergence is possible (the compromise): speaking the language of the other.
We will globally
comment on our proposal of isomorphism: the attractions will correspond to
the representation that the speakers make on the reference groups. Each language is
situated on an extreme -as we have seen on table 2 so, these will be the two poles
that attract the speakers to one side or the other. In this frame, the second fact, the
one of warp will correspond to the social representation of language B (this means,
the way how the speakers conceive it positively) that will make the attraction incline to
the side of this language. And finally, will also act as the compromising fact (or
"butterfly") that will be the positive representation of the social net from its
second language and will influence upon the intention of convergence with the speakers of
isomorphism that we have arrived to lets us delimit:
1. -The number of
variables that take part in the replacement process.
characteristics of these variables.
3. -New aspects:
Realise that there is another language, that is not A or B (7).
consideration of four type of speakers: monolingual in L1, bilingual +L1 that L2,
bilingual L1 / L2, bilingual +L2 that L1, monolingual L2.
5. - The reference
group as a central settlement of the process of linguistic replacement structure
that coincides with the theory of social representations.
proposal of a theoretical model to study the uses of languages (dependent variable)
related to the following independent variables: the social representations of the
language, the groups of reference and the social nets. Let us see in the following table.
Table 4. Relationship
between the independent variables and the dependant variables (the use of languages)
Source: Own elaboration.
between all these facts are part of the hypothesis that we aimed to corroborate. Take note
that the table wants to show us is the self-alimentation between all the variables.
As we have already
seen, in this research we are comparing the evolution of the results from the
investigation of 1993, with the ones just obtained in October 2000. The method used for
the research has been the inquiry, mainly because we wanted to compare the results that we
arrived to with the concept of linguistic attitude and these were always quantitative.
The sample has been in both
occasions representative from the students around 16 years old from all over Catalonia.
(Second of BUP at 1993, and first of batxillerat in 2000), from public to private
schooling. Within this field, the sample was divided depending on the type of
agglomeration (Barcelona/suburbs/urban/half rural), depending on the social extraction,
and depending on the linguistic orientation of the centre. The details have been picked up
from the same High schools, where we had already given the questionnaire out in 1993. And
they have been taken from the 432 inquiries obtained in 1993, and the 443 from the year
2000, where the reliability has always been very high: they have surpassed 0.9767 in both
Referring to the
treatment of statistics, we realise it from the habitual packet of the social science
investigation, the Statistical Package of Social Science (SPSS) in its 8.0
version (registered software). We apply to the results of fieldwork the analysis of
frequencies and the percentage of each variable (bi-variable analysis) and the
multivariable analysis: Discriminate analysis and multiple regression. We also use a new
method of induction techniques by graphs, SIPINA: Interactive System for the Processes of
Investigation Non-Ramificated proposed by Rakotomalala (1997
and 2000), that we will comment at point Nr. 4.2.2.