Hasil untuk "Discourse analysis"

Suharto Suharto, Hasriani Hasriani, Ardillah Abu et al.

Religious moderation in Indonesia faces a paradox, on the one hand, it is intensively promoted through state policies and public discourse; on the other hand, practices of intolerance and ideological polarization continue to emerge at the community level. The dominance of normative, top-down approaches in the literature tends to position moderation as a regulatory agenda, while religious social spaces such as mosques where everyday meanings and religious authority are actively produced remain relatively underexplored. Consequently, a theoretical gap persists in understanding how religious moderation is constructed organically through communitarian communication practices. This study examines the communication strategies of the Gerakan Subuh Berkah (GSB), initiated by the Dewan Masjid Indonesia of Palu City, as a mosque-based intervention model for strengthening religious moderation. Employing a qualitative case study design, data were collected through in-depth interviews, participant observation, and document analysis. The findings indicate that GSB has developed a dialogical and inclusive da‘wah ecosystem that integrates congregational worship, persuasive religious education, and participatory social activities. This strategy reinforces social capital (bonding, bridging, and linking), contributing to enhanced social cohesion and the mitigation of radicalization risks. The study underscores that religious moderation is not only effectively advanced through policy instruments but also through sustainable and participatory community-based religious communication. Keywords: Dakwah communication, religious moderation, Gerakan Subuh Berkah, social  capital, community resilience   ABSTRAK Moderasi beragama di Indonesia menghadapi paradox, di satu sisi dipromosikan secara intensif melalui kebijakan negara dan wacana publik, namun di sisi lain praktik intoleransi dan polarisasi ideologis tetap muncul pada level komunitas. Dominasi pendekatan normatif-top down dalam literatur cenderung memposisikan moderasi sebagai agenda regulatif, sementara ruang-ruang sosial keagamaan seperti masjid yang justru menjadi arena produksi makna dan otoritas keagamaan sehari-hari relatif kurang dikaji secara mendalam. Akibatnya, terdapat kesenjangan teoretis dalam memahami bagaimana moderasi beragama dibangun secara organik melalui praktik komunikasi komunitarian. Penelitian ini mengkaji strategi komunikasi Gerakan Subuh Berkah (GSB) yang digagas oleh Dewan Masjid Indonesia Kota Palu sebagai model intervensi berbasis masjid dalam penguatan moderasi beragama. Dengan desain studi kasus kualitatif, data diperoleh melalui wawancara mendalam, observasi partisipan, dan analisis dokumen. Hasil penelitian menunjukkan bahwa GSB membangun ekosistem dakwah dialogis dan inklusif yang mengintegrasikan ibadah berjemaah, edukasi persuasif, dan aktivitas sosial partisipatif. Strategi ini memperkuat modal sosial (bonding, bridging, linking) yang berdampak pada peningkatan kohesi sosial dan mitigasi risiko radikalisasi. Studi ini menegaskan bahwa moderasi beragama tidak hanya efektif melalui instrumen kebijakan, tetapi juga melalui komunikasi dakwah berbasis komunitas yang berkelanjutan dan partisipatif. Kata Kunci: Komunikasi dakwah, moderasi beragama, Gerakan Subuh Berkah, modal sosial, ketahanan masyarakat

Bo Shui, Xinran Zhu

Asynchronous, text-based discourse-such as students' posts in discussion forums-is widely used to support collaborative learning. However, the distributed and evolving nature of such discourse often makes it difficult to see how ideas connect, develop, and build on one another over time. As a result, learners may struggle to recognize relationships among ideas-a process that is critical for idea advancement in productive collaborative discourse. To address this challenge, we explore how large language models (LLMs) can provide representational guidance by modeling student discourse as a Knowledge Synthesis Graph (KSG). The KSG identifies ideas from student discourse and visualizes their epistemic relationships, externalizing the current state of collaborative knowledge in a form that can support further inquiry and idea advancement. In this study, we present the design of the KSG and evaluate the LLM-based approach for constructing KSGs from authentic student discourse data. Through multi-round human-expert coding and prompt iteration, our results demonstrate the feasibility of using our approach to construct reliable KSGs across different models. This work provides a technical foundation for modeling collaborative discourse with LLMs and offers pedagogical implications for augmenting complex knowledge work in collaborative learning environments.

Harry Collins, Simon Thorne

We explore the capabilities of Large Language Models (LLMs) by comparing the way they gather data with the way humans build knowledge. Here we examine how scientific knowledge is made and compare it with LLMs. The argument is structured by reference to two figures, one representing scientific knowledge and the other LLMs. In a 2014 study, scientists explain how they choose to ignore a 'fringe science' paper in the domain of gravitational wave physics: the decisions are made largely as a result of tacit knowledge built up in social discourse, most spoken discourse, within closed groups of experts. It is argued that LLMs cannot or do not currently access such discourse, but it is typical of the early formation of scientific knowledge. LLMs 'understanding' builds on written literatures and is therefore insecure in the case of the initial stages of knowledge building. We refer to Colin Fraser's 'Dumb Monty Hall problem' where in 2023 ChatGPT failed though a year later or so later LLMs were succeeding. We argue that this is not a matter of improvement in LLMs ability to reason but in the change in the body of human written discourse on which they can draw (or changes being put in by humans 'by hand'). We then invent a new Monty Hall prompt and compare the responses of a panel of LLMs and a panel of humans: they are starkly different but we explain that the previous mechanisms will soon allow the LLMs to align themselves to humans once more. Finally, we look at 'overshadowing' where a settled body of discourse becomes so dominant that LLMs fail to respond to small variations in prompts which render the old answers nonsensical. The 'intelligence' we argue is in the humans not the LLMs

Elizabeth Kyomugisha

Background Polygamy remains a significant marital institution in Uganda, where multilingualism intersects with family structure, gender roles, and social status. While legal and health implications of polygamous unions have been studied, little attention has been paid to the role of language in shaping intra-household dynamics. This study investigates how linguistic diversity within Ugandan polygamous marriages reflects and constructs social hierarchies, emotional relationships, and marital strategies. Methods This qualitative research analyzes 200 anonymized micro-narratives collected between 2023 and 2024 via the SIDINL Newsletters platform, spanning 20 Ugandan languages. Narrative analysis, sociolinguistic analysis, and discourse analysis were employed to examine language use in polygamous households. Stories were gathered from husbands, co-wives, children, and in-laws, transcribed, translated, and coded for themes such as hierarchy, emotional tone, and language switching. Results Linguistic choices, such as honorifics, metaphors, and strategic code-switching, consistently signaled status, emotional positioning, and power dynamics within polygamous families. Multilingualism was used not only to navigate household conflict and intimacy but also as a tool for forming polygamous unions, particularly in urban, border, or migratory contexts. Regional storytelling patterns revealed culturally specific framings of polygamy: moral and spiritual in the West, inheritance-based in the North, pragmatic in the East, and justice-oriented in the West Nile. Selective multilingualism emerged as a strategy where language fluency influenced marriage decisions and access to social capital. Conclusions Language functions as both a social instrument and a strategic resource in Ugandan polygamous marriages. It actively shapes family hierarchies, emotional dynamics, and mobility. Recognizing the role of language in household organization is essential for understanding marital structures in multilingual societies. These findings highlight the need for culturally grounded, linguistically sensitive approaches in both research and policy design.

Lanni Bu, Lauren Levine, Amir Zeldes

Recent LLM benchmarks have tested models on a range of phenomena, but are still focused primarily on natural language understanding for extraction of explicit information, such as QA or summarization, with responses often targeting information from individual sentences. We are still lacking more challenging, and importantly also multilingual, benchmarks focusing on implicit information and pragmatic inferences across larger documents in the context of discourse tracking: integrating and aggregating information across sentences, paragraphs and multiple speaker utterances. To this end, we present DiscoTrack, an LLM benchmark targeting a range of tasks across 12 languages and four levels of discourse understanding: salience recognition, entity tracking, discourse relations and bridging inference. Our evaluation shows that these tasks remain challenging, even for state-of-the-art models.

Huiyao Chen, Yi Yang, Yinghui Li et al.

Existing long-document question answering systems typically process texts as flat sequences or use heuristic chunking, which overlook the discourse structures that naturally guide human comprehension. We present a discourse-aware hierarchical framework that leverages rhetorical structure theory (RST) for long document question answering. Our approach converts discourse trees into sentence-level representations and employs LLM-enhanced node representations to bridge structural and semantic information. The framework involves three key innovations: language-universal discourse parsing for lengthy documents, LLM-based enhancement of discourse relation nodes, and structure-guided hierarchical retrieval. Extensive experiments on four datasets demonstrate consistent improvements over existing approaches through the incorporation of discourse structure, across multiple genres and languages. Moreover, the proposed framework exhibits strong robustness across diverse document types and linguistic settings.

Cristina BLEORŢU, Miguel CUEVAS ALONSO

-

Natalia Rodriguez Blanco

The present article is concerned with the multilingual news coverage from Agence-France Presse (AFP) about the South American country of Bolivia. Firstly, the theoretical and methodological approaches are outlined in order to characterise the plurality of contexts giving rise to AFP’s coverage of the Bolivian 2020 general elections. Secondly, an analysis is proposed that contrasts these multilingual versions in terms of framing devices and translation shifts, aiming at exploring the ways in which media stakeholders represent the Bolivian reality. Thirdly, the findings of this analysis are contextualised with reference to a cross-linguistic comparison of newspaper corpora. When comparing the Spanish, French, and English versions, the first two are found to be more aligned at the level of discourse patterns. The ultimate purpose of this case-study is to observe the presence of translation in plurilingual news settings, where the role of translators often goes unacknowledged within

Victor Andres Ayma Quirita, Juan David Cardenas, Walter Aliaga et al.

Over the years, technology has rapidly evolved, with social networks and media emerging as prominent examples. Social media facilitates the creation and exchange of content, while social networks connect individuals and organizations based on shared interests or values. Among these platforms, Twitter (now X) has gained significant attention and has been a valuable tool for monitoring reputation and brands, including political discourse. With the widespread use of social networks, the demand for advanced textual data analysis techniques, particularly sentiment analysis, has grown. Sentiment analysis, or polarity classification, aims to gauge the positivity or negativity of textual content, providing information on public opinion dynamics. In the context of the Peruvian presidential elections of 2021, characterized by political uncertainty exacerbated by the COVID-19 pandemic, social media platforms, especially Twitter/X, emerged as vital arenas for political discourse and participation. The election featured various candidates representing a wide spectrum of ideologies, reflecting the complexities of Peruvian society. Against this backdrop, this research uses Twitter/X data to analyze public sentiment toward presidential candidates and election-related topics. Through state-of-the-art sentiment analysis algorithms, this study categorizes tweets into positive, negative, or neutral sentiments, revealing key trends and themes driving public opinion on Twitter/X during the election period. By examining sentiment dynamics in response to major events and comparing Twitter/X sentiment trends with official voting data, the research aims to assess the predictive power of social media sentiment analysis in forecasting electoral outcomes. Therefore, this research contributes to understanding the role of social media in shaping political discourse and public opinion during elections. By conducting sentiment analysis on Twitter data, valuable insights are offered to policymakers, political analysts, and researchers who want to leverage social media to understand and forecast electoral trends.

Alina Landowska, Maciej Skorski, Krzysztof Rajda

People worldwide are currently confronted with a number of technological challenges, which act as a potent source of uncertainty. The uncertainty arising from the volatility and unpredictability of technology (such as AI) and its potential consequences is widely discussed on social media. This study uses BERTopic modelling along with sentiment and emotion analysis on 1.5 million tweets from 2021 to 2023 to identify anticipated tech-driven futures and capture the emotions communicated by 400 key opinion leaders (KOLs). Findings indicate positive sentiment significantly outweighs negative, with a prevailing dominance of positive anticipatory emotions. Specifically, the 'Hope' score is approximately 10.33\% higher than the median 'Anxiety' score. KOLs emphasize 'Optimism' and benefits over 'Pessimism' and challenges. The study emphasizes the important role KOLs play in shaping future visions through anticipatory discourse and emotional tone during times of technological uncertainty.

Kai-Robin Lange, Jonas Rieger, Niklas Benner et al.

From a monarchy to a democracy, to a dictatorship and back to a democracy -- the German political landscape has been constantly changing ever since the first German national state was formed in 1871. After World War II, the Federal Republic of Germany was formed in 1949. Since then every plenary session of the German Bundestag was logged and even has been digitized over the course of the last few years. We analyze these texts using a time series variant of the topic model LDA to investigate which events had a lasting effect on the political discourse and how the political topics changed over time. This allows us to detect changes in word frequency (and thus key discussion points) in political discourse.

F. B. Lazaridou, A. Heinz

Introduction Evidence shows that racism can have a negative effect on mental health in the lived experiences of Black people and People of Colour. In critical theory discourse including postcolonial and decolonial approaches, racism is suggested to be an everyday phenomenon. Additionally, racism specifically targets the perceived cultural and phenotypic foreignness of Black migrants and migrants Of Colour, as well as the ascribed migrant status attributed to the perceived foreignness of racialized persons who do not actually have any direct migration experiences. Objectives The stigma associated with severe mental disorders such as psychosis has historically been applied to Black people and People of Colour who have been engaged in anti-racist activism as a form of punishment and social control. Higher incidence rates of psychosis in racialized communities have frequently been conceptualized as cultural differences in family composition and levels of expressed emotion in families. The objective of this study is to sensitively investigate psychosis as a potential mental health consequence of racism. Methods The incidence rates of psychosis - positive symptoms, negative symptoms, non-affective psychosis disorders and first episode psychosis - among migrants by country of migration were compiled in an umbrella review, which offers a summary of meta-analyses. Quantitative research has the limitation of enabling the observation of patterns but not allowing an understanding of the reasons behind them to be theorized through the data. Therefore, qualitative methods complement the quantitative data. Twenty people of diverse genders who self-identified as Black people or People of Colour in Berlin were interviewed about their experiences of racism and sexism and about how those experiences affected their mental health. Results The umbrella review found an association between migration and psychosis, with migration from the Caribbean and African countries showing the strongest correlation. A constant comparative analysis of the qualitative data suggests that racism contributes to the emergence of a subclinical psychosis symptomatology profile that consists of a sense of differentness, negative self-awareness, paranoid ideation regarding general persecution, and self-questioning with self-esteem instability. Conclusions The findings are interpreted as a situational diagnosis, as coined by the psychiatrist and political philosopher Frantz Fanon in the seminal book ‘Black Skin, White Masks’ (1975). The findings are also contextualized within a critique of institutional racism, both historically and currently, and within an intersectional discussion of the need for structural competency and the provision of safety for racialized groups in clinical settings. Disclosure of Interest None Declared

Mariya Kozolup, Olha Patiyevych, Halyna Kryzhanivska et al.

Background. Nowadays, almost all indexed journals expect submissions in English, which is a great challenge for exophonic authors. Code-switching context, where cross-language effects, especially native language interference, are well distinct, is critical for approaching the dilemma. Navigating the complicated issues of language-related challenges will be impossible without referring to three crucial levels of written production: lexical, syntactic, and textual. In our investigation, we address the nature of potential errors and their inter- and intralingual origins. In particular, we identify and interpret the deviations from Standard English in scholarly research writing of Ukrainian authors in the field of life sciences, exemplify and classify errors into categories based on the type of language misuse. Materials and Methods. Language material for the study comprised 50 manuscripts submitted by authors from Ukraine to the journal “Studia Biologica”. This research is a mixed-method study encompassing descriptive qualitative and descriptive quantitative methods. Content analysis was employed as the data gathering technique. The analysis of texts was focused on tracing deviations from consistent principles and rules of Standard English and linguistic features of English research discourse and encompassed such steps as highlighting the error, cross-checking and stating the deviation, listing and classifying the errors, and tracing a possible connection of the error to authors’ first language interference. Results. The study identified language areas where Ukrainian authors fail to effectively communicate their ideas to the global academic community. At the textual level, the problem areas encompass defective paragraph structure and excessive verbosity. At the syntactic level, the most critical deviations from the language and stylistic norm comprised misuses of word order and clauses, wordy and confusing sentences with multiple issues that hinder the readability of text. The most widespread grammatical mistakes include missing predicates, faulty subject-verb agreement, incorrect forms of the verb, and inappropriate use of articles, pronouns, demonstratives and quantifiers. At the lexical level, the prevalent errors relate to various types of loan translation, but also include improper word choices and poor vocabulary. Orthographic mistakes, though in minority, refer to the spelling of toponyms, capitalisation, switching from American to British orthographic standards and other random spelling errors. Conclusions. An insight into the nature of the analysed deviations suggests the presence of both intra- and interlingual factors that cause mistakes in papers submitted for publication in the field of life sciences. The error analysis can be beneficial in the educational process for both educators and practitioners. Proper understanding of the functional mechanism of the mistakes might increase the awareness of the potential pitfalls and consequently help avoid them. The classification of errors can be adopted in the educational process and contribute to the development of error pedagogy.

Ramos Asafo-Adjei, Francis Bukari, Ernest Kwasi Klu

This instrumental qualitative case study examined the acceptance speech of Professor Naana Jane Opoku-Agyemang, the first female vice presidential candidate of a major political party in Ghana. The study aimed to identify the discourse elements and cohesive devices used in the speech and to evaluate how she conveyed her political ideology and gender identity. Specifically, the research question was: which discourse elements and cohesive devices were employed, and how were political ideology and gender identity manifested in the speech? The data consisted of an eight-page speech delivered on July 28, 2020, in Accra, containing 3,972 words. Fairclough's (2015) Three-Dimensional textual analysis model was used for data analysis, which includes description, presentation and interpretation, and explanation of social reasons. Results showed that Professor Opoku-Agyemang used discourse elements and cohesive devices such as 'however,' 'also,' and 'second' to highlight her academic and political expertise and present herself as a qualified vice presidential candidate. She aimed to inspire women and encourage them to aspire to any position, including becoming the president of Ghana. The study's findings contribute to existing literature and can inform future research on female political representation, exposing the approaches of a marginalized group and their use of speeches to gain recognition within Ghana’s political space and beyond.

Patrick Huber, Giuseppe Carenini

With a growing need for robust and general discourse structures in many downstream tasks and real-world applications, the current lack of high-quality, high-quantity discourse trees poses a severe shortcoming. In order the alleviate this limitation, we propose a new strategy to generate tree structures in a task-agnostic, unsupervised fashion by extending a latent tree induction framework with an auto-encoding objective. The proposed approach can be applied to any tree-structured objective, such as syntactic parsing, discourse parsing and others. However, due to the especially difficult annotation process to generate discourse trees, we initially develop such method to complement task-specific models in generating much larger and more diverse discourse treebanks.

Salim Hafid, Sebastian Schellhammer, Sandra Bringay et al.

Scientific topics, claims and resources are increasingly debated as part of online discourse, where prominent examples include discourse related to COVID-19 or climate change. This has led to both significant societal impact and increased interest in scientific online discourse from various disciplines. For instance, communication studies aim at a deeper understanding of biases, quality or spreading pattern of scientific information whereas computational methods have been proposed to extract, classify or verify scientific claims using NLP and IR techniques. However, research across disciplines currently suffers from both a lack of robust definitions of the various forms of science-relatedness as well as appropriate ground truth data for distinguishing them. In this work, we contribute (a) an annotation framework and corresponding definitions for different forms of scientific relatedness of online discourse in Tweets, (b) an expert-annotated dataset of 1261 tweets obtained through our labeling framework reaching an average Fleiss Kappa $κ$ of 0.63, (c) a multi-label classifier trained on our data able to detect science-relatedness with 89% F1 and also able to detect distinct forms of scientific knowledge (claims, references). With this work we aim to lay the foundation for developing and evaluating robust methods for analysing science as part of large-scale online discourse.

Shulin Zhang, Jixing Li, John Hale

Pro-drop is commonly seen in many languages, but its discourse motivations have not been well characterized. Inspired by the topic chain theory in Chinese, this study shows how character-verb usage continuity distinguishes dropped pronouns from overt references to story characters. We model the choice to drop vs. not drop as a function of character-verb continuity. The results show that omitted subjects have higher character history-current verb continuity salience than non-omitted subjects. This is consistent with the idea that discourse coherence with a particular topic, such as a story character, indeed facilitates the omission of pronouns in languages and contexts where they are optional.

Tatiana Silva Tavares, Kênia Lara Silva, Regina Garcia de Lima et al.

ABSTRACT Objective: To analyze the experiences of families in the exercise of the rights of children with chronic conditions in public health, education and social assistance institutions. Method: ethnographic multiple case study, with qualitative approach, following the theoretical approach of Boaventura Santos. Experiences of the families of these children in a city were studied through interviews with family members, managers and professionals from social institutions (35), participant observations in social spaces (13) and creation of eco-maps (3). Critical Discourse Analysis was performed. Results: the offer of services is lower than the demand, and exclusion processes persist. Given the hegemony of neoliberal and normality ideologies, meetings between family members and professionals revealed obstacles to civil rights; however, when these ideologies were challenged, the realization of their rights was enhanced. Final considerations: the care to promote civil rights requires family members, managers and professionals to develop subjectivities that overcome neoliberal and normality ideologies, recognizing these children as subjects of law.

Wanqiu Long, Xinyi Cai, James E. M. Reid et al.

Text corpora annotated with language-related properties are an important resource for the development of Language Technology. The current work contributes a new resource for Chinese Language Technology and for Chinese-English translation, in the form of a set of TED talks (some originally given in English, some in Chinese) that have been annotated with discourse relations in the style of the Penn Discourse TreeBank, adapted to properties of Chinese text that are not present in English. The resource is currently unique in annotating discourse-level properties of planned spoken monologues rather than of written text. An inter-annotator agreement study demonstrates that the annotation scheme is able to achieve highly reliable results.

Patrick Huber, Giuseppe Carenini

Discourse parsing could not yet take full advantage of the neural NLP revolution, mostly due to the lack of annotated datasets. We propose a novel approach that uses distant supervision on an auxiliary task (sentiment classification), to generate abundant data for RST-style discourse structure prediction. Our approach combines a neural variant of multiple-instance learning, using document-level supervision, with an optimal CKY-style tree generation algorithm. In a series of experiments, we train a discourse parser (for only structure prediction) on our automatically generated dataset and compare it with parsers trained on human-annotated corpora (news domain RST-DT and Instructional domain). Results indicate that while our parser does not yet match the performance of a parser trained and tested on the same dataset (intra-domain), it does perform remarkably well on the much more difficult and arguably more useful task of inter-domain discourse structure prediction, where the parser is trained on one domain and tested/applied on another one.