Real-naturalism in literature from the perspective of Distant Reading: a study of the corpus OBras

Abstract

This article presents a methodology with computational resources for the investigation of natural language processing. Added to this factor, we analyzed aspects of the construction of meaning of words said to be symptomatic, as Domício Proença Filho (1995) points out, in realistic and naturalist works of Brazilian literature. Such literary productions are available in the OBras corpus, which is composed of a vast set of narratives in the public domain (LINGUATECA, 2008). In this sense, we observe what the before-mentioned literary critic affirms when analyzing the novel O mulato. In the scholar's conception, the nouns, verbs and adjectives in this work are symptomatic words, since they create, in the reader, a suggestive ambience. This observation serves for the critic to make a survey about Brazilian real-naturalism, in which, according to Proença Filho, the authors conveyed the negative aspects of society in their works (PROENÇA FILHO, 1995). Therefore, in this study, a theoretical basis based on the Digital Humanities was sought, with a focus on what has been called Distant Reading (MORETTI, 2008). Digital Humanities concern the transversality of areas of knowledge and of social communitarian representations within the digital electronic space (PEREIRA, 2015) and distant reading concerns an attributed concept that proposed the computation of literary data. In addition to Moretti's (2008) criticism, our work is based on the meanings of Underwood (2017), Wolfreys (2009), Bastos (2019), Almeida (2019), Alves (2019), Rocha (2019), Ogiboski ( 2012), Higuchi (2021), Silva (2021), Santos (2019) and Hockey (2018).