Publication: Constructicografia multilíngue da flexão verbal : uma abordagem construcionista contrastiva para tempo, modo e aspecto em português e espanhol
Authors
Duarte Marção, Natália
item.page.secondaryauthor
Escuela Internacional de Doctorado
item.page.director
Timponi Torrent, Tiago ; Valenzuela Manzanares, Javier
Publisher
Universidad de Murcia
publication.page.editor
publication.page.department
DOI
item.page.type
info:eu-repo/semantics/doctoralThesis
Description
Abstract
Esta tesis presenta el modelado de las construcciones de tiempo, aspecto y modo verbal (TAM) en portugués brasileño y en español desde una perspectiva construccionista contrastiva. El objetivo es contribuir tanto al desarrollo del Constructicon de FrameNet Brasil (TORRENT ET AL., 2018) como a la iniciativa MoCCA – Model of Comparative Concepts for Aligning Constructicons (ALMEIDA ET AL., 2024). El trabajo se basa en fundamentos teóricos de la Lingüística Cognitiva, tales como la Semántica de Marcos (FILLMORE, 1982) y la Gramática de Construcciones de Berkeley (KAY & FILLMORE, 1999), y adopta la metodología de análisis de la FrameNet Brasil. Además, incorpora los Comparative Concepts propuestos por Croft (2022) como base para el análisis contrastivo. La metodología combina: (a) un levantamiento del tratamiento que reciben las construcciones TAM en ambas lenguas, con base en estudios descriptivos y contrastivos presentes en compendios gramaticales y materiales didácticos; (b) la identificación de correspondencias de traducción entre construcciones TAM en corpus bilingües alineados; y (c) la modelización computacional de dichas construcciones. Dado que fue necesario delimitar el alcance del estudio para adecuarse al tiempo disponible en un programa de doctorado, la tesis se enfoca en las construcciones TAM del modo indicativo, incluyendo tanto los tiempos simples como los tiempos compuestos. En total, se modelaron 53 construcciones: 27 en portugués y 26 en español. En el caso del portugués, se modelaron 8 construcciones TAM y 19 construcciones adicionales necesarias para la correcta definición de los elementos constituyentes. En el caso del español, se modelaron 7 construcciones TAM y 19 construcciones adicionales. El análisis contrastivo sigue la metodología propuesta por Laviola et al. (en prensa), que calcula la similitud del coseno entre las construcciones modeladas. Esta métrica se complementa con un estudio de corpus contrastivo, destinado a verificar si las diferencias de alcance formal y funcional entre construcciones del mismo tipo en cada lengua, tal como fueron definidas en la modelización, se correlacionan con un mayor o menor grado de correspondencia entre ellas. Los resultados indican que la asociación de los Comparative Concepts a la modelización permite identificar de forma efectiva similitudes y discrepancias entre las construcciones TAM en portugués y español. Asimismo, los datos revelan que un estudio contrastivo de corpus entre lenguas distintas debe tener en cuenta múltiples factores contextuales, ya que estos son fundamentales para interpretar adecuadamente los niveles de similitud o, en particular, de discrepancia entre las construcciones.
This work presents the modeling of tense, aspect, and mood (TAM) constructions in Brazilian Portuguese and Spanish from a contrastive constructionist perspective. The goal is to contribute to both the development of the FrameNet Brasil Constructicon (TORRENT ET AL., 2018) and the MoCCA initiative — Model of Comparative Concepts for Aligning Constructicons (ALMEIDA ET AL., 2024). Grounded in the theoretical framework of Cognitive Linguistics, the study draws on Frame Semantics (FILLMORE, 1982) and Berkeley Construction Grammar (KAY & FILLMORE, 1999), following the analytical methodology adopted by FrameNet Brasil. Furthermore, the research incorporates Croft’s (2022) notion of Comparative Concepts to inform its contrastive approach. The methodology combines: (a) a review of how TAM constructions are addressed in descriptive and contrastive studies found in reference grammars and pedagogical materials for both languages; (b) the identification of translation correspondences between TAM constructions using aligned bilingual corpora; and (c) the computational modeling of these constructions. Due to the need to delimit the scope of the research to fit within the timeframe of a doctoral program, the study focuses exclusively on indicative mood TAM constructions, encompassing both simple tense inflectional forms and auxiliary tense constructions. In total, 53 constructions were modeled — 27 in Portuguese and 26 in Spanish. For Portuguese, 8 TAM constructions and 19 additional constructions (necessary to define the components of the TAM structures) were modeled. For Spanish, the modeling included 7 TAM constructions and 19 additional constructions. The contrastive analysis applies the methodology proposed by Laviola et al. (forthcoming), which calculates cosine similarity between the modeled constructions. This quantitative approach is supplemented by a corpus-based contrastive study aimed at investigating whether formal and functional distinctions among constructions of the same type in each language, as defined in the modeling, correlate with stronger or weaker cross-linguistic correspondences. Findings suggest that the incorporation of Comparative Concepts into the modeling process yields promising results in capturing both similarities and divergences between TAM constructions in Portuguese and Spanish. Moreover, the data indicate that corpus-based contrastive studies must account for a range of contextual factors, as these are essential to accurately interpret the degree of similarity or, more critically, the divergences between constructions across languages.
This work presents the modeling of tense, aspect, and mood (TAM) constructions in Brazilian Portuguese and Spanish from a contrastive constructionist perspective. The goal is to contribute to both the development of the FrameNet Brasil Constructicon (TORRENT ET AL., 2018) and the MoCCA initiative — Model of Comparative Concepts for Aligning Constructicons (ALMEIDA ET AL., 2024). Grounded in the theoretical framework of Cognitive Linguistics, the study draws on Frame Semantics (FILLMORE, 1982) and Berkeley Construction Grammar (KAY & FILLMORE, 1999), following the analytical methodology adopted by FrameNet Brasil. Furthermore, the research incorporates Croft’s (2022) notion of Comparative Concepts to inform its contrastive approach. The methodology combines: (a) a review of how TAM constructions are addressed in descriptive and contrastive studies found in reference grammars and pedagogical materials for both languages; (b) the identification of translation correspondences between TAM constructions using aligned bilingual corpora; and (c) the computational modeling of these constructions. Due to the need to delimit the scope of the research to fit within the timeframe of a doctoral program, the study focuses exclusively on indicative mood TAM constructions, encompassing both simple tense inflectional forms and auxiliary tense constructions. In total, 53 constructions were modeled — 27 in Portuguese and 26 in Spanish. For Portuguese, 8 TAM constructions and 19 additional constructions (necessary to define the components of the TAM structures) were modeled. For Spanish, the modeling included 7 TAM constructions and 19 additional constructions. The contrastive analysis applies the methodology proposed by Laviola et al. (forthcoming), which calculates cosine similarity between the modeled constructions. This quantitative approach is supplemented by a corpus-based contrastive study aimed at investigating whether formal and functional distinctions among constructions of the same type in each language, as defined in the modeling, correlate with stronger or weaker cross-linguistic correspondences. Findings suggest that the incorporation of Comparative Concepts into the modeling process yields promising results in capturing both similarities and divergences between TAM constructions in Portuguese and Spanish. Moreover, the data indicate that corpus-based contrastive studies must account for a range of contextual factors, as these are essential to accurately interpret the degree of similarity or, more critically, the divergences between constructions across languages.
publication.page.subject
Citation
item.page.embargo
Collections
Ir a Estadísticas
Este ítem está sujeto a una licencia Creative Commons. http://creativecommons.org/licenses/by-nc-nd/4.0/