Busca de medida de complexidade textual numa abordagem funcionalista

Abstract

Since the end of the 1970s, the search for measures of comprehension, intelligibility and complexity of texts has been interesting, mainly, to scholars from several areas, to institutions of teaching evaluation and to Language teachers in Brazil. Many formulas and techniques based on structural features of the text came up, besides the subjective evaluation based on age and personal preferences. However, it is unknown a textual complexity measure based on verbal and visual aspects in a functionalist approach. This article reports an investigation that aims to search an instrument of evaluation of the level of complexity of verbal and visual texts by using computational technological resources, based on categories and functionalist principles. For that, it has, above all, Functional Linguistics and Linguistics of Corpus as theoretical and methodological orientation for the analysis of the verbal and visual resources. It takes as study sample texts and tests of the National High School Exam (ENEM), submitted to a quantitative and statistical comparison with data of reference texts of the Corpus Lácio-Ref (ALUÍSIO et al., 2003). The quantitative comparison between corpora has provided data for the elaboration of formulas of measures that evaluate, in a scalar way, Portuguese language texts, from three levels of complexity: structural, cognitive and visual. It is concluded that the proposal is of easy access and that by using computational technological resources, it can be provided the analysis of the complexity of only one text as well as of texts in large amounts.