A emergência de construções resultativas de verbos sobre violência no português brasileiro

Abstract

This thesis aims to describe the cognitive semantic emergence of transitive and in-transitive resultative constructions following a cognitive-semantic approach, in order to cover its productivity and distribution in the discursive-mediatic violence context. The hypothesis is that emergence occurs because of general order cognitive processes, such as metonymy (BARCELONA, 2003; KOVECSES; RADDEN, 1998), probabilistic attraction between collocations and constructions – collostructions – (STEFANOWITCH; GRIES, 2003) and Idealized-Cognitive-Model (ICM) categorization (LAKOFF, 1987), among other reasons. Traditionally, causative verbal alternation is about an intransitive construction deriving from a transitive one and its verb still keeps the same meaning of “cause” and “change of state” (RAPPAPORT HOVAV; LEVIN, 2012; RAPPAPORT HOVAV, 2014). Opposite to the syntactic transformational derivation, Goldberg (1995) understands causative constructions are interdependent and they build a productive relation through inheritance links to meet semanticpragmatic conditions and purposes. Therefore, they are named after transitive and intransitive resultative constructions, since both forms highlight the [EFFECT] explicitly. This study chose its own terminology for the phenomenon, calling them in-transitive constructions. Those which emerged only in the transitive form were treated as transitive-only. The research uses descriptive methodology and is based on hypothetical-deductive procedures, collecting the data from the CASS corpus on violence news, published by Brazilian media between 2014-2015, and made available by the Lancaster University (UK). From an almost-two-million-word corpus, 6,988 concordance lines were extracted with change-of-state verbs associated to the VIOLENCE concept. The transitive-only resultative constructions were analyzed apart from the in-transitive ones, considering the Semantic Coherence and Correspondence principles (GOLDBERG, 1995). The first group has emerged most of the analyzed constructions, and a metonymic chain has been verified: SUBJETCT FOR AGENT - AGENT FOR CAUSE - SUBJECT FOR CAUSE; as well as DIRECT OBJECT FOR PATIENT - PATIENT FOR EFFECT – DIRECT OBJECT FOR EFFECT. Regarding the transitive constructions without the [direct object] argument, the CAUSE-FOR-EFFECT and the PROCESS-FOR-EFFECT metonymies have participated effectively in these constructions’ emergences, which were statistically accepted as collostructions that showed the discursive purpose of generally communicating the [EFFECT] of the [PROCESS] represented by the construction verb. On the in-transitive constructions, some verbs, as “disparar” (fire/shoot) and “explodir” (explode), have emerged in both transitive and intransitive forms, while others came up only in the transitive form. However, they remained in the same group due to the possibility of emerging in both variants in other discourses and corpora. Yet the CAUSE-FOR-EFFECT and the EFFECT-FOR-CAUSE metonymies have contributed with the intransitive variants. The [EFFECT] dominance over the [CAUSE] was verified, since the first was able to profile the arguments/participants [direct object/patient] for the transitive constructions and the [subject/patient] for the intransitive ones. The latter came up as collostructions for pointing their collocational exclusivity. Most of the constructions in the research have emerged as transitiveresultative and some even got restricted to this form, which revealed CAUSALITY/VIOLENCE ICM is built around the [CAUSE]-[PROCESS]-[EFFECT] conceptual interaction, attracting the transitive variant to the category prototypicity.