174 Rosanna Sornicola and Silvia Calamai
especially in Europe. The individual speaker is considered as a source and, con-
sequently, all his/her characteristics should be documented: in this sense, data
collection should be anthropological and microlinguistic.^2
As far as the principle of microscopy is concerned, sound archives enable us
to observe the opposition between ‘local’ and ‘global’ at its best: microscopic study
allows the best correlation between the three dimensions of variation (diaphasy,
diatopy, diastraty). Sometimes, it is precisely microscopy (the study of the oscil-
lations determined by a certain phenomenon in the texts of a group of speak-
ers of a community) which allows the understanding of linguistic change, while
macroscopy helps us understand, on a large scale, the spreading factors external
to the phenomenon. The difficulty lies, once again, in relating microscopic and
macroscopic dimensions.
The principle of oscillation and that of linguistic context put two closely
connected characteristics together under the same label. The oscillations in the
realization of a certain phenomenon (in relation with Mathesius’ concept of
‘potentiality’) directly recall the concept of ‘variable’ and its range of variants. As
is well-known, linguistic context represents a fundamental source of variation:
context-induced allophonic dispersion, for example, characterizes synchronic
variability and variation and can prelude diachronic changes.
The principle of having recourse to spontaneous speech assumes that precisely
at the spontaneous speech level the speaker selects variants which are in his/her
competence, but do not come out in formal style or in response to direct ques-
tions. Direct elicitation methods, in fact, do not ensure the attainment of those
levels of ‘automatic’ spontaneity which appear to be crucial when studying stylistic
variation. Even within identical portions of spontaneous speech, differences in the
level of self-control have been detected in which the speaker provides oscillations
between Italianized and dialectal forms.
Several authors belonging to the European tradition have concentrated on
the importance of psychological and linguistic differences between speakers. In
Gilliéron, Dauzat, Gardette and Duraffour (e.g., Gillierón & Roques 1912; Gillieron
1918, 1921; Dauzat 1900, 1922; Duraffour 1932; Gardette 1983) it is possible to iden-
tify a French line of theoretical and methodological reflection which gives promi-
nence to the psychological aspects of production and knowledge of patois (Sornicola
2002). This certainly constitutes a crucial problem, especially for sociophonetic- The archive thus has the individual speaker as its basic unit of analysis. The archiving proce-
dure is carried out by the interviewer him/herself (in order to minimize the loss of information
from the person who was present at the recorded event and the person who archives it). The
interviewer also prepares a sheet for every speaker in which to note down both the objective
features of traditional sociolinguistic research (sex, age, education, job) and the subjective fea-
tures (personality, attitudes, motivations that emerged during the interview).