Moliere's Raisonneurs: a quantitative study of distinctive linguistic patterns

Francesca Frontini (1, 2) , Mohamed Amine Boukhaled (2) , Jean-Gabriel Ganascia (2)

1 CNR-ILC - Istituto di Linguistica Computazionale "Antonio Zampolli"
2 ACASA - Agents Cognitifs et Apprentissage Symbolique Automatique

Great authors of plays and novels are often renowned for the ability to create memorable characters that take on a life of their own and become almost as real as living persons to their readers/audience. The study of characterization, that is, of how it is that authors manage to achieve this, has become a well-researched topic in corpus stylistics: for instance (Mahlberg, 2012) attempts to identify typical lexical patterns for memorable characters in the work of Dickens by extracting those lexical bundles that stand out (namely are over-represented) in comparison to a general corpus. In other works, authorship attribution methods are applied to the different characters of a play to identify whether the author has been able to give each of them with a “distinct” voice. For instance (Vogel and Lynch, 2008) compare the dialogues of individual characters in a Shakespeare play against the rest of the play or even against all plays in the Shakespearean corpus. In Frontini et al (2015), we propose a methodology for the extraction of significant patterns that enables literary critics to verify the degree of characterization of each character with respect to the others and to automatically induce a list of linguistic features that are significant and representative for that character. The proposed methodology relies on sequential data mining for the extraction of linguistic patterns and on correspondence analysis for the comparison of pattern frequencies in each character and the visual representation of such differences. We chose to apply this analysis to Moliere’s plays and the protagonists of those plays. In this work we focus on the figure of the raisonneurs, characters who take part in discussions with comical protagonists providing a counterpart to their follies. Such characters were interpreted at times as spokesmen for Moliere himself, and the voice of reason, at other times as comical characters themselves and no less foolish than their opponents. Hawcroft’s essay Reasoning with fools (2007) highlights the differences between five of these characters based on their role in the plot. Using this analysis as guidance, we compare significant linguistic patterns in order to see how these differences are marked by the author. We do this by adapting the discourse traits of each of them to the communicative function they needs to fulfill (Biber and Conrad 2009).

Mots clés: computational stylistics; Molière; correspondence analysis; syntactic pattern extraction

Domaines: Linguistique; Informatique [cs]

Francesca Frontini, Mohamed Amine Boukhaled, Jean-Gabriel Ganascia. Moliere's Raisonneurs: a quantitative study of distinctive linguistic patterns. Corpus Linguistics 2015, Jul 2015, Lancaster, United Kingdom. ⟨hal-01202089⟩