Wikipedia stands as a remarkable field for linguistic study. The quantity and diversity of accessible textual data types such as articles,
various talk pages, and revision histories (including precise traceability of contributors and modifications) enable the investigation of
several axes of linguistic variation using quantitative analysis. In this pilot study, we examine three axes of variation: thematic variation,
diachronic variation, and variation according to text type. Based on a corpus consisting of 108 French articles and their associated talk
pages within Wikipedia, we observe significant differences on each of these axes for a panel of linguistic indicators, such as indicators
of size, lexical diversity and distribution of parts of speech. This constitute an initial step within a larger project, providing a foundation
for interesting reflections to subsequently investigate the articulation between individual and collective levels of linguistic variation
within a given community.