
A Pilot Corpus Study on Microdiachronic Complexity Features in Brazilian Labor Law
Abstract
This microdiachronic study focuses on the Consolidação das Leis do Trabalho (CLT: Decreto-Lei nº 5.452/1943), the Brazilian labor law, extracted from the LEX-BR-Ius corpus (Ferrari & Marques, 2022)—a collection of Brazilian federal statutory laws in force (754 distinct laws, totaling 3,296,993 tokens). The diachronic annotation in Modest XML (Hardie, 2014) signals, among other features, the amendments to each law. In the pilot study, UDPipe was used to map nine distinct complexity dimensions, including word order variation, nominalization, subordination, passive constructions, and lexical density. A Principal Component Analysis (PCA) was conducted on year-level syntactic complexity measures, and the resulting complexity scores were compared across four historical periods. Results indicate substantial stability in individual feature frequencies, but a marked increase in the PCA-derived composite index of syntactic complexity in the most recent period. This suggests a process of register complexification characterized by greater overall syntactic densification, rather than by sharp increases in individual constructions considered in isolation.
The study can be replicated and deepened, as both the corpus and the scripts are available as open data.
© 2026 Lúcia de Almeida Ferrari, Luciana Dias de Macedo, published by Ubiquity Press
This work is licensed under the Creative Commons Attribution 4.0 License.