Doutorado em Estudos Linguísticos
URI Permanente para esta coleção
Nível: Doutorado
Ano de início: 2016
Conceito atual na CAPES: 4
Ato normativo: Homologado pelo CNE (Portaria MEC Nº 609, de 14/03/2019).
Publicação no DOU 18 de março de 2019, seç. 1, p.135 - Parecer CNE/CES nº 487/2018, Processo no 23001.000335/2018-51).
Periodicidade de seleção: Anual
Área(s) de concentração: Teoria e Análise Linguística
Url do curso: https://linguistica.ufes.br/pt-br/pos-graduacao/PPGEL/detalhes-do-curso?id=1511
Navegar
Submissões Recentes
- ItemEthos, autoralidade e identidade discursivas em cenografias do ciberespaço(Universidade Federal do Espírito Santo, 2025-06-30) Miranda, André Freitas; Nascimento, Jarbas Vargas; https://orcid.org/0000-0002-2002-1752; http://lattes.cnpq.br/3171836873464711; https://orcid.org/0000-0002-9596-2178; http://lattes.cnpq.br/9973800091776101; Vidon, Luciano Novaes; https://orcid.org/0000-0003-2231-6279; http://lattes.cnpq.br/3136561352267602; Alves Junior, Mário Acrísio; https://orcid.org/0000-0002-1732-8251; http://lattes.cnpq.br/8555982390613118 ; Cano, Márcio Rogério de Oliveira; https://orcid.org/0000-0001-5502-6951; http://lattes.cnpq.br/4320952131815734; Ferreira, Anderson; https://orcid.org/0000-0001-7980-5773; http://lattes.cnpq.br/1006738639592351Cette thèse s’inscrit dans le champ de l’Analyse du Discours de tradition française (AD) et se concentre sur la construction de l’autoralité (auteur), de l’identité et de l’ethos discursifs dans les scénographies numériques du cyberespace, à partir de l’activité du journal A Gazeta sur son compte X/Twitter. Le corpus comprend des publications liées à des nouvelles, reportages, caricatures et articles d’opinion publiés sur le site officiel du journal, organisées en cinq moments énonciatifs : février et avril 2021, février 2022 et février 2023. Nous partons de la question suivante : celui qui énonce un discours dans une condition de production donnée peut-il être considéré comme son auteur, même lorsque ce discours est approprié par diverses formations discursives et circule sur différentes plateformes ? À partir de cette problématique, nous examinons comment le journal construit son image auctoriale et identitaire dans la fluidité et la performativité du langage numérique. Il s’agit de réfléchir à l’autorat dans l’environnement numérique et à ses implications pour les études du discours. Dans ce contexte, nous comprenons que la production de sens, la constitution des sujets et la construction de l’image de l’énonciateur s’opèrent à travers des pratiques propres au numérique et une négociation constante des ethos, de l’identité et du positionnement. Nos analyses s’appuient sur un cadre théorique composé principalement de Maingueneau (2010 ; 2015 ; 2019), Paveau (2021), Charaudeau (2001 ; 2008 ; 2014), Bakhtine (2009 ; 2017), Foucault (2013) et Agamben (2007), avec un accent particulier sur la notion d’autoralité comme geste discursif. L’énonciation numérique mobilise une complétude énonciative qui construit à la fois une identité institutionnelle et une image d’auteur liée au journal, révélant différents niveaux de responsabilité discursive. Nous cherchons ainsi à comprendre comment la présence numérique institutionnelle du sujet projette des images de soi dans le discours, construisant des ethos discursifs multiples, qui s’articulent, se confrontent et se négocient constamment dans le cyberespace à travers l’hypermédia. Nous constatons que l’autoralité est un processus qui émerge comme fonction et geste dans différentes scènes d’énonciation. A Gazeta, dans ce processus, occupe des positions auctoriales multiples et projette des images de soi qui se construisent, se tendent et se renégocient dans l’espace hypermédiatique.
- ItemHipercorreção e mobilização de significados sociais: percepções sobre o uso do relativo qual em um estudo open-guise(Universidade Federal do Espírito Santo, 2025-03-28) Santos, André Poltronieri; Tesch, Leila Maria; https://orcid.org/0000-0003-3919-1230; http://lattes.cnpq.br/9705222558363890; https://orcid.org/0000-0002-4905-1242; http://lattes.cnpq.br/6125651735094828 ; Chaves, Raquel Gomes; https://orcid.org/0000-0002-6310-0194; http://lattes.cnpq.br/2546227696082610; Damasceno, Gesieny Laurett Neves; https://orcid.org/0000-0001-6556-9968; http://lattes.cnpq.br/1635569588468079 ; Oushiro, Livia; https://orcid.org/0000-0002-2165-3305; http://lattes.cnpq.br/9719372821079096; Mendes, Ronald Beline; https://orcid.org/0000-0002-1510-7180; http://lattes.cnpq.br/0651144091367586This dissertation aims to investigate the perceptions of listeners/readers on hypercorrect uses of the relative pronoun qual using an open-guise test (Soukup, 2013; 2016). We justify the interest in these uses because the relative pronoun qual is rarely used in speech and writing, but exhibits a relatively high frequency of hypercorrections, as demonstrated by Lorevice (2012) and Silva, A. (2021). The main hypotesis of this research is that perceptions of hypercorrections would be conditioned by the kind of relative clause (prepositional phrase chopping or a hypercorrect use of qual), as well as the participants’ academic background. The experiment’s protocol consisted of presenting two pairs of stimuli, recorded by the same speaker, which differed only in terms of the relative pronouns (que, in PP-chopping, and a hypercorrect use of qual). Even though we used the voices of two different subjects, each participant listened to the stimuli in the voice of only one of them. In addition to the spoken stimuli, each participant evaluated a written stimulus that presented i) two PP-chopping relative clauses, or ii) two hypercorrections of qual. The results of the experiment show that, even though participants knew they were listening to, and evaluating the voices of the same speaker, and that the audio recordings were simulations, i. e., not samples of spontaneous speech, they showed distinct perceptions about the two variants presented on the stimuli. These perceptions were also conditioned to some extent to social variables, but structural factors showed stronger effects on the perceptions, particularly the fact that participants had heard the second pair of stimuli, which presented a similar contrast between the variants of the first pair, suggestive of a priming effect actuation (Greenbaum, 1976; Teixeira, 2016). Overall, the analysis of regression models and several paired t-tests showed that participants perceived the stimuli with hypercorrection of qual as having higher status and being more pedantic. However, we observed that participants’ academic background was an important conditioner in their perceptions of admiration: while the group of Letras (Languages and Literature) and related courses had perceptions of less admiration about the stimuli with the hypercorrection of qual of the second pair, participants of non-related courses did not differentiate such stimuli in terms of this characteristic. As for the written stimuli, despite similarities between the two groups of participants, only those from Letras and related courses perceived the hypercorrections of qual as more pedantic. In short, these results allow us to argue that the hypercorrect use of the relative qual convey social meanings.
- ItemDa solidariedade à “casa da ãe Joana”: análise do discurso digital sobre os refugiados no Facebook(Universidade Federal do Espírito Santo, 2025-05-29) Hora, Lidia Gurgel Neves; Lima, Fábio Luiz Malini de; https://orcid.org/0000-0002-2405-9109; http://lattes.cnpq.br/8284809605215682; https://orcid.org/0000-0002-8346-9968; http://lattes.cnpq.br/8401529474263427; Almeida, Julia Maria Costa de; http://orcid.org/0000-0003-2430-4116; http://lattes.cnpq.br/3648370916765504; Tomazi, Micheline Mattedi; https://orcid.org/0000-0002-2246-7061; http://lattes.cnpq.br/4783716565631781; Baronas, Roberto Leiser; https://orcid.org/0000-0003-0758-0370; http://lattes.cnpq.br/4613001301744682; Soares, Rosana de Lima; https://orcid.org/0000-0003-4250-9537; http://lattes.cnpq.br/5241011640369563This study employed a qualitative-quantitative method to analyze discourses concerning refugees on Facebook during 2020 and 2021, the initial phase of the COVID-19 pandemic. By integrating Data Science — specifically, the Perspectivist Method of Social Network Analysis — and Digital Discourse Analysis, the research examined 38,410 posts and 2.8 million interactions. The connections between words and actors were graphically represented as discursive perspectives. From the word-based perspectives, key themes emerged, including solidarity with refugees, the Brazilian government’s Welcoming Operation, and vulnerabilities associated with gender, race, and poverty. The actor-based perspective revealed collaborations between institutions and influencers that amplified the visibility of refugee mobilization and social initiatives. Notably, Brazilian far-right military and political figures appeared unified in their support for refugee reception — a stance contrasting with European xenophobia. This divergence may be attributed to the strategic alignment of such actions with anti-communist ideologies concerning Venezuela and the persecution of Christians in Afghanistan. An analysis of commercial media perspectives presented limited coverage of refugee issues, whereas alternative media outlets actively denounced human rights abuses and potentially hazardous political alliances. In both word and actor-based perspectives, the prominence of international organizations was evident. Within user comments, despite prevailing sentiments of solidarity, instances of hate speech and xenophobia were identified, often framed around concerns regarding refugee numbers, public safety, and health risks. These discourses frequently aligned with the global far-right. The study underscores the necessity for enhanced social media regulation, including stricter comment moderation and strategies to counter hate speech in digital interactions. It also highlights the importance of discourse analysis in capturing Brazil’s sociopolitical specificities, and advocates for the systematic archiving and processing of the country’s linguistic data. Furthermore, the findings suggest that discourse analysis could play a pivotal role in mitigating biases within AI-driven language models.
- ItemA construção discursiva do feminicídio em notícias jornalísticas: uma análise sistêmico-funcional(Universidade Federal do Espírito Santo, 2025-04-11) Santos Neto, Alfredo Evangelista dos; Neves, Gesieny Laurett Damasceno; https://orcid.org/0000-0001-6556-9968; http://lattes.cnpq.br/1635569588468079; http://lattes.cnpq.br/7853775277291538; Simões, Alex Caldas; https://orcid.org/0000-0002-6661-6436; http://lattes.cnpq.br/2757065561066716; Baroni, Gabriela Do Couto; http://lattes.cnpq.br/6756007512619656; Machado, Marcos Roberto; http://lattes.cnpq.br/4733094588152083; Rodrigues, Violeta Virginia; https://orcid.org/0000-0003-1789-5675; http://lattes.cnpq.br/7907063278349571Systemic Functional Linguistics (SFL) stands out for its approach that views language as a set of interconnected systems, in which grammatical structures are analyzed in relation to the meanings they convey and the functions they perform in texts. Based on this perspective, this study investigates how femicide is discursively constructed in news reports published on online news portals, adopting a functionalist approach centered on SFL. The analysis is grounded in the theoretical foundations of Linguistic Functionalism (Givón, 1984, 1990, 1995, 2001) and SFL, as proposed by Halliday (1994), Halliday and Matthiessen (2004, 2014), Thompson (2014), Martin and White (2005), Martin and Rose (2008), and Hasan (1985), in addition to the descriptive proposal of Material Processes developed by Damasceno (2016). The general objective is to examine how the components of the textual, interpersonal, and ideational metafunctions are encoded in news reports on femicide, as well as to understand the meanings that emerge from these codifications. Specifically, the research proposes four lines of analysis: (i) identifying Theme, Rheme, and their meaning effects (textual metafunction); (ii) observing interpersonal relations between clausal subjects (interpersonal metafunction); (iii) analyzing the Processes, Participants, and Circumstances involved in the actions (ideational metafunction); and (iv) mapping and comparing discursive patterns through the use of Self-Organizing Maps (SOM), a statistical technique based on unsupervised neural networks. The methodological approach employs thirteen analytical parameters that encompass pragmatic and discursive properties related to the three metafunctions. Among these parameters, the following stand out: Semantic Representation of the Theme (RSE), related to the textual metafunction; Informational Status of the process responsible (EIS), linked to the interpersonal metafunction; and Position of Circumstances (PCI), associated with the ideational metafunction. The study also incorporates extralinguistic aspects by comparing the patterns identified across the five Brazilian regions (Central-West, North, Northeast, South, and Southeast), with particular emphasis on the contrastive analysis between the state with the highest femicide rate (Mato Grosso do Sul) and the one with the lowest (Ceará). The corpus comprises 33 news articles reporting on 11 distinct cases of femicide, totaling 634 clauses with Material Processes—375 of which are active and 259 passive. The application of the SOM network, combined with Pearson’s correlation, enabled the identification of twelve recurring linguistic patterns (six in active clauses and six in passive ones) and nine statistically significant correlations. Among these, the most prominent are: (i) the correlation between RSE and EIS in passive clauses, revealing a tendency toward explicit subjects when the process occupies the thematic position; (ii) the association between the number and position of Circumstances in active clauses, with a preference for initial and final placement; and (iii) the relationship between the Thematic Role of the Subject and the Type of Material Doing, emphasizing the responsibility attributed to the agent. The results indicate that the distribution of linguistic patterns in femicide news reports reflects intentional discursive choices that shape how these cases are represented and socially interpreted. Such grammatical choices—as in the inclusion of the circumstance “in front of several company employees,” an example from the research corpus—not only locate the action but also add interpretative value, intensifying the perception of the crime’s brutality and public exposure. Thus, linguistic organization contributes to the construction of meaning and social valuation of femicide.
- ItemEfeitos de sentido coconstruídos em fake news e comentários no ex-twitter: tecnorreferenciação, tecnointertextualidade e patemização(Universidade Federal do Espírito Santo, 2024-12-06) Franco, Kátia Regina; Souza Júnior, Rivaldo Capistrano de; https://orcid.org/0000-0002-3731-7613; http://lattes.cnpq.br/7433150215859023; https://orcid.org/0009-0002-4713-3081; http://lattes.cnpq.br/6895152511671476; Oliveira, Mônica Lopes Smiderle de; Costa, Ilioni Augusta da; Lins, Maria da Penha Pereira; Marchon, Amanda HeiderichThis thesis aims to investigate how the processes of referentiation and intertextuality are mobilized in fake news produced on X, formerly Twitter. The theoretical foundation supporting the research is based on the categories of Textual Linguistics, specifically referentiation and intertextuality (Mondada, 2001; Mondada & Dubois, 2003; Cavalcante, 2003, 2004, 2011, 2016; Cavalcante et al., 2016, 2017, 2019, 2022; Elias & Cavalcante, 2017), the Argumentative Discourse Theory on polemical argumentation (Amossy, 2005, 2008, 2011, 2017, 2018), and Digital Discourse Analysis (Paveau, 2021), which describes the characteristics of technodiscourse. The theoretical framework also includes the discursive treatment of emotion (Charaudeau, 2007; Elgin, 2007; Brun & Kuenzle, 2008) and the characterization of fake news (Wardle, 2017; Wardle & Derakhshan, 2017; Shul et al.). Through an ecological analysis (Paveau, 2021), we seek to understand how the processes of referentiation and intertextuality are employed in the technodiscursive production of fake news on the social network X, creating effects of pathos and credibility. Our hypothesis is that the production of technotexts encompasses many additional elements that were dispensable in pre-digital writing, due to the inherent characteristics of digital text, such as expansion, composite composition, nonlinearization, relationality, investigability, and unpredictability, which influence the meaning effects of texts activated by referentiation and intertextuality, potentially updating polemics and provoking pathos effects. To test our hypothesis, we constructed a corpus of 26 texts, composed of 5 posts verified as fake news, 13 comments, and 8 tweets that revisit the discourse object focused on in the selected fake news, redocumented through an advanced search of the keyword on the X platform. The results indicate that the characteristics of technogenres, especially expansion, nonlinearization, and relationality, articulated with the discursive and linguistic gestures enabled by the social network X, such as interaction buttons and the insertion of images, can (re)update or even redefine the notion of textualization in the digital era and, consequently, the observation of the processes of referentiation, ntertextuality, and argumentation. The categories selected for analysis have their linguistic processes impacted by technical gestures, as often, the discourse object is identified and/or referenced in the sequence of tweets or comments; intertextuality is frequently established in comment-shares and through lexical or imagistic links; and all these enunciative gestures cooperate to update the polemic, the argumentation, and the adherence to the viewpoint of the speaker/disseminator of fake news, according to the pathos bias effects constructed in the interaction.