The purpose of this paper is to analyze if implementing the hermeneutically created negation scope can improve the performance of Sentiment Analysis tools when applied to João Cabral de Melo Neto's poems. To achieve this goal, the paper proposes a hybrid approach that combines lexical analysis based on two dictionaries with a scope of negation created from the author's own poems. Furthermore, some algorithms were used to confirm the negation in the context of Cabralina's poetry. The experiment showed a significant improvement in performance scores when applying one of the dictionaries associated with the scope of negation. This underscores the significance of computational linguists' endeavors to create dictionaries for literary texts and reinforces the need for collaboration between areas of knowledge.