Questões de responsabilidade enunciativa e autoria em texto acadêmico produzido com auxílio do ChatGPT
Domínios de Lingu@gem · 27.08.2024 · 18 · e1838
DOI: 10.14393/dlv18a2024-38 ↗Kurz gefasst
Fachartikel in der Zeitschrift Domínios de Lingu@gem, online erschienen am 27.08.2024. Verfasst von Rosangela Alves dos Santos Bernardino und 1 weiteren Autor:innen. Die Arbeit wurde bisher 2-mal zitiert (Crossref) und nennt 19 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY-NC-ND 4.0).
Abstract (Originalsprache)
This paper proposes the study of enunciative responsibility based on the analysis of an academic text produced with the help of ChatGPT, seeking to problematize the attribution of authorship in this context of human interaction with an Artificial Intelligence tool. The methodology involves carrying out an experiment with ChatGPT, with the corpus being a cross-section of ongoing research in the Institutional Scientific Initiation Scholarship Program (PIBIC), in the 2024-2025 edition. The work follows a qualitative, descriptive-interpretive approach and focuses on the enunciative level of textual analysis, as proposed by Adam (2011). Theoretically, it is based on Textual Discourse Analysis, in dialog with the assumptions of the Bakhtin Circle and the enunciative-pragmatic point of view approach. The results show that ChatGPT inserts the voices of others into the materiality of the text, indicating the imputation of points of view (POS), but also emits a certain “critical” position, according to the commands provided. The expressive linguistic marks of imputation were the mediative markers, such as “according to”, “as”, “such as”, and the types of speech representation, direct and indirect speech. Enunciative engagement was signaled through modal expressions, in which an appreciative, complimentary tone was observed. As for the stylistic-compositional elements of the text, we highlight the use of syntactic structures, phrasing and “plastered” lexical choices that seem to constitute a mechanical pattern of writing. The data calls into question authorship, since the text does not express the figure of an author as a subject at the origin of their own POS.
Zitieren
APA
Bernardino, R. A. D. S., & Guerra, Z. P. (2024). Questões de responsabilidade enunciativa e autoria em texto acadêmico produzido com auxílio do ChatGPT. Domínios de Lingu@gem, 18, e1838. https://doi.org/10.14393/dlv18a2024-38
Vancouver
Bernardino RADS, Guerra ZP. Questões de responsabilidade enunciativa e autoria em texto acadêmico produzido com auxílio do ChatGPT. Domínios de Lingu@gem. 2024;18:e1838. doi:10.14393/dlv18a2024-38
BibTeX
@article{Bernardino2024,
author = {Bernardino, Rosangela Alves dos Santos and Guerra, Zailton Pinheiro},
title = {Questões de responsabilidade enunciativa e autoria em texto acadêmico produzido com auxílio do ChatGPT},
journal = {Domínios de Lingu@gem},
year = {2024},
volume = {18},
pages = {e1838},
doi = {10.14393/dlv18a2024-38}
}
EUVIMED