What Can Be Done With The Results Of Varbrul?

Abstract

This work aims at bringing up some discussion about the analysis of statistical data, and a priori conclusions, highlighting the importance of considering, among other issues, the orthogonality in the distribution of data, since some reflections were generated by some of the results obtained with the VARBRUL program. Therefore, we will present the (i) results from the analysis of tu/você switch in the localities of VARSUL where the process exists, and (ii) analysis of Noun Agreement (NA) in the Paraná cities - Londrina, Irati and Pato Branco. In the analysis of NA, we realized that there are differences regarding the selection of variables, depending on the type of colonization of the analyzed region (for Londrina and Pato Branco, internal colonization; for Irati, external colonization) - the reason why Curitiba city, the "old" Paraná, was not considered in this paper. In the analysis of tu/você switch (LOREGIAN-PENKAL, 2004), we emphasize the importance of considering the analysis of the variation in the individual which, mutatis mutandis, can be applied to individual results of the NA in the three cities. The data analyzes followed the quantitative sociolinguistics steps with VARBRUL package, and took into account extralinguistic and linguistic factors. We hope to demonstrate how the VARSUL database has been contributing to the description of the variable phenomena in Brazilian Portuguese.