510 JOURNAL OF LAW AND POLICY
Hypothesis 2 stated that an individual’s idiolectal style would
stay relatively stable despite the course of time. In order to
confirm this hypothesis, samples from the same individual in
MT1 and MT2 were compared with each other. In Figures 5–7,
points in the graph indicate an intraspeaker/writer IIS result,
i.e., an IIS value after comparing samples from the same
subjects in two separate points in their lives.
Results show that this second hypothesis is confirmed for
both the phonological and the morphosyntactic modules. Figures
5 and 6 illustrate results in these two modules for the three
languages. As can be seen, IIS results for all the modules range
between 0.8 and 0.9, which is high, as expected, since 1 on the
IIS continuum means maximum similarity.
With regard to the discourse-pragmatic modules, hypothesis
2 could only be tested for the Catalan and English modules,
since the Spanish corpus for this module did not contain data in
two measurement times. Hypothesis 2 is also confirmed with all
three methods of Catalan and English. With method 3 (Figure
7), all IIS values are quite high, as expected, with the majority
ranging between 0.9 and 0.7.
Figure 4: Interspeaker IIS results for the discourse-pragmatic modules.