| dc.description.abstract | This paper aims to analyze the conceptual variation of legal domain terms, seen here as a procedure that contributes to the implementation of a legal domain ontology. This procedure is an exploratory and interdisciplinary study based on the Communicative Theory of Terminology (Cabré), dealing with the concept of terminological variation. As a methodology, the Lexicology of Verticality (Wichter) is used to allow for the variation analysis through conceptual categories. Femicide, procuring, and embezzlement were the terms examined in the study, three crimes set forth in the Penal Code of Brazil. For this analysis, we used three text genres for each term with different levels of specialty: one legislation, one specialized article, and one press report. As the work unfolded, conceptual categories that emerged from the analyzed texts and that marked off the comparative work were used. The following conceptual categories were proposed: Legislation, Classification, Type, Subject of Crime, Motivation, Characteristics, Aggravating Circumstances, Penalty, and related Crimes. Some of them presented internal features, which were called subcategories: Legislation (law), Legislation (article of the Penal Code of Brazil), Legislation (popular name), Subject of Crime (active), Subject of Crime (passive), Motivation (individual), Motivation (collective), Penalty (time), Penalty (regime), Penalty (with aggravating circumstances), Penalty (if wrongful act) and Penalty (previous). The analysis was divided into two parts, being the first organized per term and the second per text genre. In the first stage (per term) we identified which linguistic realization favors the target-audience of the text genre at hand. We also highlighted different levels of conceptual depth in the classification of crimes. In the second stage (contrastive analysis among the text genres), the least specialized text presented more conceptual voids, and the most specialized one presented less conceptual voids. Three cases of conceptual emptying were also identified, where the conceptual density decreases from the most specialized to the least specialized texts. We further observed that the specialized article more repeatedly presented new conceptual contents in the text. Finally, we presented the structural components that are prototypical of each text genre for the conceptual variation of crimes. These findings contribute to the implementation of a legal domain ontology in several aspects: we realized that through the analysis of different text genres it is possible to identify the cause and effect relation of crime, its characteristics. It is also possible to define what crime is not, which different laws influence penalty, that crime classification directly influences the length of penalty and how it is to be served, and so forth. In sum, these texts allowed us to recover concept attributes that can be applied both to entity creation and to the entity proper positioning in a conceptual structure, assigning them with properties and setting different types of relations. We reached the conclusion that the terms present different conceptual densities, varying according to the discursive situation engendered in the text genre and to the level of specialty held by the audience the text is aimed at. | en |
