Dicendi Verbs as Indicators of Coherence in Abstracts

a human and automated analysis

Keywords: Semantic Coherence, Word Semantics, Dicendi Verbs, Large Language Model

Abstract

The school summary plays a fundamental role in developing the ability to synthesize and understand texts, promoting the interpretation and the organized expression of the contents of the source text. This study investigates how the choice of dicendi verbs (verbs of saying) in summaries can reflect the coherence and intention of the source text. Given the importance of coherence for the quality of summaries, the research explores the potential of large-scale language models (LLMs) in identifying or suggesting dicendi verbs that are coherent with the purpose of the text, contributing to a more accurate automatic evaluation. The experiments reported in this article used summaries proposed for a given text in a college entrance exam, on which experts and LLMs suggested appropriate verbs. The suggestions were compared and evaluated for coherence with the source text. The results indicate that, although LLMs are capable of identifying appropriate verbs in some contexts, they have limitations when compared to human interpretation. We conclude that LLMs can serve as auxiliary tools in education, despite the challenges related to accuracy and context in automated linguistic assessment.

Author Biographies

Osmar de Oliveira Braz Junior, Universidade do Estado de Santa Catarina

Osmar de Oliveira Braz Junior holds a degree in Computer Science from the University of Southern Santa Catarina (1997) and a master's degree in Production Engineering from the Federal University of Santa Catarina (2000). He is currently an assistant professor at the State University of Santa Catarina (UDESC) and an hourly professor at the University of Southern Santa Catarina (UNISUL). He has experience in the area of ​​Computer Science, with an emphasis on Software Engineering, working mainly in the following areas: information systems, distance education, software engineering and database.

Ana Julia Araujo Sanchuki, Universidade Tecnológica Federal do Paraná

Undergraduate student of the Bachelor's degree course at UTFPR; currently, she develops her research work on the relationship between AI and the teaching of Portuguese Language, focusing on the school summary genre.

Roberlei Alves Bertucci, Universidade Tecnológica Federal do Paraná

Roberlei Alves Bertucci holds a degree in Portuguese-English Literature from PUCPR (2004); a master's degree in Literature (Linguistic Studies) from UFPR (2007) and a PhD in Linguistics from USP (2011). He completed part of his doctorate at Université Paris 8 (2009-2010). He conducted postdoctoral research at Bar-Ilan University in Israel (2012). He is currently a professor at the Federal Technological University of Paraná (UTFPR). He is interested in different grammatical (formal) processes of meaning production in natural languages, such as: syntax, semantics and pragmatics of natural languages, especially Brazilian Portuguese; linguistic description and analysis in the verbal and nominal domains, especially through technological tools; and the application of linguistic foundations and discoveries to digital technological tools.

Renato Fileto, Universidade Federal de Santa Catarina

Renato Fileto holds a Bachelor degree in Computer Science from the Federal University of Uberlândia (1992), a Master degree (1994) and a Doctorate degree (2003) in Computer Science from Campinas State University, Brazil, with an internship at Georgia Institute of Technology, USA (2002), and a Post-Doctorate from the University of São Paulo (2012). His research carrier has been intertwined with activities in the industry. Since 2006, he is a permanent professor at the Department of Informatics and Statistics (INE) of Santa Catarina Federal University (UFSC), in Florianópolis-SC, Brazil. His research area is data science, with the focus in intelligent systems for data analytics.

Published
2025-07-01
How to Cite
de Oliveira Braz Junior, O., Araujo Sanchuki, A. J., Alves Bertucci, R., & Fileto, R. (2025). Dicendi Verbs as Indicators of Coherence in Abstracts: a human and automated analysis. Linguamática, 17(1), preprint. Retrieved from https://linguamatica.com/index.php/linguamatica/article/view/461
Issue
Vol. 17 No. 1
Section
Research Articles

Copyright (c) 2025 Osmar de Oliveira Braz Junior, Ana Julia Araujo Sanchuki, Roberlei Alves Bertucci, Renato Fileto

Creative Commons License

This work is licensed under a Creative Commons Attribution 4.0 International License.

Authors who publish with this journal agree to the following terms:

    1. Authors retain copyright and grant the journal right of first publication with the work simultaneously licensed under a Creative Commons Attribution License that allows others to share the work with an acknowledgement of the work's authorship and initial publication in this journal.
    2. Authors are able to enter into separate, additional contractual arrangements for the non-exclusive distribution of the journal's published version of the work (e.g., post it to an institutional repository or publish it in a book), with an acknowledgement of its initial publication in this journal.
    3. Authors are permitted and encouraged to post their work online (e.g., in institutional repositories or on their website) prior to and during the submission process, as it can lead to productive exchanges, as well as earlier and greater citation of published work (See The Effect of Open Access).