In this paper, the political conflict in Venezuela in 2019 is approached from a corpus linguistic point of view. The conflict between Juan Guaidó, the speaker of the parliament, and Nicolás Maduro, who won the internationally unrecognized 2018 presidential election, escalated on January 23, 2019, when Guaidó proclaimed himself the legitimate president of Venezuela. By comparatively analyzing the tweets of the two politicians three months before and after January 23, 2019, a corpus-based discourse analysis will be conducted to investigate whether and to what extent linguistic patterns (especially most frequent words and their co-occurrences, as well as n-grams) change within the respective social media communication of these political opponents. The analysis reveals changes in linguistic patterns, especially with respect to co-occurrences and n-grams, detected in the corpus data, and demonstrates that politicians use Twitter to present themselves, in the case of Guaidó, as the representative of the people wanting to lead Venezuela into a democratic future, and, in the case of Maduro, as the only legitimate president and defender of Venezuela against internal and external threats. Research data is published on Zenodo.