Papers reviewed and determined not to be word norm studies. Use the flag icon to report errors or suggest re-inclusion.
18625 papers
Replication package for "CEI: A Benchmark for Evaluating Pragmatic Reasoning in Language Models" (DMLR 2026). The Contextual Emotional Inference (CEI) Benchmark is a dataset of 300 expert-authored scenarios for evaluating how well language models interpret pragmatically complex utterances in social contexts. Each scenario presents a communicative exchange involving indirect speech (sarcasm, mixed signals, strategic politeness, passive aggression, or deflection) where the speaker's literal words diverge from their actual emotional state. Three trained annotators independently labeled every scenario using Plutchik's 8 basic emotions and Valence-Arousal-Dominance ratings. This archive contains: • data/human-gold/ — 5 merged annotation CSVs (300 scenarios, 3 annotators each) • scripts/ — Pipeline, analysis, and HuggingFace upload scripts • config/ — Model definitions and pricing configuration • papers/dmlr2026/ — Paper source (LaTeX), bibliography, and figures • reports/dmlr2026/ — Baseline results (JSON) and LaTeX tables • LICENSE (MIT for code) and README.md The dataset is released under CC-BY-4.0. Code is released under MIT. GitHub: https://github.com/jon-chun/cei-tom-dataset-base HuggingFace: https://huggingface.co/datasets/jonc/cei-benchmark
Large language models (LLMs) perform strongly on many NLP tasks, but their ability to produce explicit linguistic structure remains unclear. We evaluate instruction-tuned LLMs on two structured prediction tasks for Standard Arabic: morphosyntactic tagging and labeled dependency parsing. Arabic provides a challenging testbed due to its rich morphology and orthographic ambiguity, which create strong morphology-syntax interactions. We compare zero-shot prompting with retrieval-based in-context learning (ICL) using examples from Arabic treebanks. Results show that prompt design and demonstration selection strongly affect performance: proprietary models approach supervised baselines for feature-level tagging and become competitive with specialized dependency parsers. In raw-text settings, tokenization remains challenging, though retrieval-based ICL improves both parsing and tokenization. Our analysis highlights which aspects of Arabic morphosyntax and syntax LLMs capture reliably and which remain difficult.
Misconceptions in physics are persistent explanatory frameworks that develop in early infancy and continue throughout primary, secondary, and university education, even among pre-service educators. Instead of being eradicated through formal education, these intuitive notions often coexist with scientific principles, resulting in a disjointed or contextually dependent comprehension. This review consolidates recent studies on the origin, endurance, and alteration of misunderstandings across several educational stages and physics disciplines, encompassing mechanics, thermal phenomena, energy, and electromagnetism. It emphasizes that misconceptions are influenced by perceptual experiences, linguistic norms, and evaluative methods that favour procedural fluency over explanatory reasoning. The paper analyzes instructional methods that facilitate conceptual change, highlighting hands-on and virtual experimentation, guided inquiry, diagnostic evaluation, and innovative digital and AI-driven technologies. The research identifies cross-cutting themes that emphasize the significance of cognitive conflict, explicit interaction with learners' concepts, and coherent conceptual advancement in curriculum. Significant study deficiencies encompass the necessity for longitudinal studies, examinations of transfer and durability, and a more profound inquiry into teacher education as a domain for conceptual advancement. The findings emphasize the importance of intentional conceptual education and comprehensive teacher preparation to foster lasting and significant comprehension in physics.
This article argues that in Amy Tan’s “Mother Tongue,” language functions not merely as a means of communication but as a site of symbolic power, social stratification, and identity formation within the framework of ethnic self-narration. The article demonstrates how the ideology of Standard English produces a hierarchy of linguistic legitimacy that systematically marginalizes speakers positioned outside its normative framework. Rather than treating Tan’s narrative as a purely individual memoir, this study reads it as a strategic articulation of collective ethnic experience. In this sense, “Mother Tongue” operates as a political act of self-narration that transforms private memory into a public critique of linguistic discrimination. By foregrounding her mother’s so-called “broken English,” Tan exposes the ideological construction of linguistic competence and the institutional inequalities embedded in the American educational system and public discourse. From this perspective, Tan’s narrative constitutes an aesthetic and political resistance that challenges the authority of standardized linguistic norms. Ultimately, “Mother Tongue” emerges as a counter-hegemonic intervention that reconfigures both individual and collective identity, recenters marginalized voices, and interrogates the symbolic mechanisms of power that operate through language. Existing scholarship has predominantly framed the essay as a narrative of identity reconciliation; this article, by contrast, argues that it stages a structural critique of linguistic authority.
Over the past six decades, a variety of systems have been developed for representing sign language forms, from Stokoe Notation (Stokoe, 1960) to SignWriting (Sutton, 1999) and lexical database schemas.Each was designed with specific goals and applications, leading to a fragmented landscape of representations.To enable greater interoperability and data sharing among sign language users and researchers, we propose a robust approach to translating between notation systems.As a first step in this direction, we introduce a formal mapping framework between HamNoSys and the SL CatForm coding schema, describe its implementation, and present empirical evidence of its performance.An extensive evaluation of mapping mismatches revealed improvements to the mapping logic needed to further advance the HNS2CF mapping tool.However, the initial version of the system already achieves an overall accuracy of 76.7% and an in-depth analysis reveals that many apparent mismatches stem from annotator disagreement rather than mapping errors, indicating that the tool's actual accuracy is even higher.These results demonstrate the feasibility and promise of establishing mapping mechanisms across sign representation systems.
Empathy and emotional regulation play a central role in the socio-affective adaptation of young adults. However, their interactions remain insufficiently documented, particularly when emotional stimuli are task-irrelevant. The present study aimed to examine the extent to which the cognitive and affective dimensions of empathy modulate emotional regulation mechanisms within a motor inhibition paradigm. Thirty-one young adults completed the Interpersonal Reactivity Index (IRI) and subsequently performed a modified Go/NoGo task in which negative, neutral, or positive images were presented as affective distractors preceding a Go or NoGo signal. Following each trial, participants rated the subjective valence of the image on a Likert scale. Results first indicated that negative images slowed response times in the Go condition, whereas positive images were associated with a reduction in error rates relative to neutral images. Furthermore, activation of inhibitory control influenced subjective emotional experience, with overall more negative valence ratings following a NoGo signal than a Go signal, particularly for positive images. Finally, analyses revealed that specific empathy dimensions, notably perspective-taking and empathic concern, were associated with more negative ratings of negative images, without exerting a robust effect on inhibitory performance. These findings suggest that inhibition is not limited to behavioral adjustment but also contributes to the implicit modulation of emotional experience, whereas empathy appears to influence subjective sensitivity to negative stimuli more than the efficiency of inhibitory control.
Music is a promising nonpharmacological intervention for enhancing healthcare outcomes. Emotional regulation is a key mechanism in this process, but results vary, possibly due to differences in music selection. This study explores the impact of AI-composed music on emotional and physiological states, as measured by the self-assessment mannequin (SAM) and heart rate variability (HRV). Healthy volunteers (n = 30) were subjected to 10 min of AI-composed music with the intention of creating positive valence and negative arousal. Before the intervention, questionnaires on personal characteristics and music preferences, and baseline HRV were measured. Subjectively, no significant change in valence was observed (Δ median 0.0 (IQR 0.0;1.0), p = 0.077), whereas arousal (Δ median -1.0 (IQR -2.0;0.0), p < 0.001) and dominance (Δ median 0.5 (IQR 0.0;1.0), p = 0.008) significantly decreased. Ordinal regression analysis revealed that a higher rating for classical music (in line with the genre of the AI-composed music) predicted a higher valence rating after the intervention (OR 2.39, 95% CI 1.16-4.92; p = 0.018). Objectively, HRV measurements decreased and suggested a less relaxed/more excited state during the music listening task compared to baseline. Subsequent analysis revealed that an increase in valence was associated with higher HRV. In conclusion, after listening to the AI-composed music, subjective arousal was significantly lower, whereas valence was overall not affected by the music. However, music preference emerged as a crucial factor influencing emotional valence, which in turn affected HRV, indicating a more relaxed state. Therefore, AI-composed music in line with individual preferences may benefit music interventions, encouraging personalized approaches in healthcare.
Gender equality remains a cornerstone of sustainable social development, yet in Zimbabwe, the persistent use of sexist language in meetings, workshops, and public gatherings continues to reinforce gender hierarchies and marginalize the girl child. This study is of critical importance as it interrogates the linguistic and cultural foundations that perpetuate gender bias, aiming to advance gender-neutral communication as a pathway to empowerment and equity. Despite numerous gender mainstreaming policies, a significant research gap exists in understanding how everyday sexist discourse sustains systemic inequalities and limits girls’ participation and agency in social, educational, and professional spaces. The primary objective of this study is to examine the prevalence, forms, and socio-cultural implications of sexist language in Zimbabwean institutional and community contexts, and to propose strategies for promoting inclusive and gender-sensitive communication. Adopting a mixed-methods design, the research combines qualitative interviews, focus group discussions, and document analysis with quantitative surveys conducted across government, educational, and civic institutions. Data will be analysed thematically and statistically to capture both the depth and breadth of linguistic gender bias. Preliminary findings indicate that sexist language is deeply embedded in traditional communication practices, institutional norms, and even policy discourse, resulting in subtle but pervasive disempowerment of girls and women. However, evidence also points to a growing awareness and readiness among educators, policymakers, and community leaders to adopt gender-neutral communication frameworks. The study’s implications are profound: by reshaping linguistic norms, Zimbabwe can cultivate environments that affirm equality, inclusivity, and respect. The research contributes to global discourse on gender and language while providing actionable recommendations for policymakers, educators, and advocacy groups to dismantle linguistic barriers and foster the full empowerment of the girl child.
Language has historically functioned as both a communicative system and a repository of cultural identity, social memory, and collective belonging. In the contemporary digital era, Artificial Intelligence (AI) technologies increasingly mediate communication practices through predictive text systems, speech recognition software, algorithmic moderation, and Large Language Models (LLMs). Although such technologies have improved efficiency, accessibility, and multilingual interaction, emerging scholarship suggests that AI systems may also reproduce existing linguistic hierarchies by privileging dominant and standardized language forms while marginalizing dialectal and non-standard varieties (Helm et al., 2024; Hofmann et al., 2024). This theoretical paper examines how AI-driven communication systems may contribute to processes of linguistic standardization and dialect marginalization within digital environments. Drawing upon Fricker’s (2007) theory of epistemic injustice, the paper introduces the concept of the Digital Language Divide to explain disparities in computational recognition, linguistic legitimacy, and technological visibility among speech communities. Unlike deterministic arguments that portray technology solely as a force of linguistic erosion, this paper adopts a balanced perspective acknowledging that AI technologies may simultaneously expand communication access while also reinforcing structural inequalities in language representation. Recent empirical studies demonstrate that some AI systems exhibit dialect prejudice, particularly toward speakers of African American English and other non-standard language varieties (Hofmann et al., 2024). Additionally, emerging scholarship on techno-linguistic bias argues that language technologies frequently prioritize dominant linguistic norms embedded within training datasets (Helm et al., 2024). Building upon these findings, this paper theorizes how algorithmic standardization may gradually influence linguistic practices, educational expectations, and communicative participation within digitally mediated societies. This paper seeks to establish a conceptual foundation for future interdisciplinary inquiry into the sociolinguistic consequences of AI-mediated communication. The study concludes by proposing directions for future empirical research, policy development, and inclusive AI design frameworks capable of supporting linguistic diversity within digital ecosystems.
Academic STEM evaluation can elicit anxiety, yet routine grading rarely captures how students semantically frame exams and wellbeing. We reconstruct these framings using behavioural forma mentis networks (BFMNs), that is, feature-rich networks of concepts linked by memory recalls and enriched with affective ratings and concreteness norms. We build BFMNs from 994 participants spanning STEM experts (N1 = 59), Italian high-schoolers (N2 = 206), physics undergraduates (N3 = 10), psychology undergraduates with math-anxiety levels (N4 = 301), and simulated students (N5 = 497) personified by a large language model (GPT-OSS 20B). Across all human groups, the concepts "exam" and "grade" were (i) perceived negatively, (ii) connected primarily to negatively valenced memory recalls, indicating a clustering of negative emotions around assessment, and (iii) framed through concepts eliciting fear and anticipation in most groups, including physics undergraduates (z-scores in the range [2.04, 2.53]). The semantic neighbourhoods of "anxiety" and "exam" overlapped three times more in human students than in GPT-based simulations, providing structural evidence of test anxiety in student populations. By contrast, experts displayed a neutral and more concrete network neighbourhood for "exam" (z = 1.87), with no clear trace of test anxiety. These negative assessment framings coexisted with positive representations of "wellbeing", which were rich in concrete associations in humans but linked to more abstract concepts in GPT digital twins. Overall, our results show that BFMNs offer a quantitative and interpretable framework to study academic anxiety and to distinguish human affective framing from current AI-based simulations
Abstract: Language holds major significance because it helps people to develop their personal identities while enabling them to pass their identities through time and it helps people to stay connected with each other in regions where international interactions take place. This article considers Kinyarwanda as a transnational linguistic system, meaning that it is not only the official language of Rwanda but it is also a culturally and communicatively binding element for the Great Lakes African region and the Rwandan diaspora. Adopting a qualitative paradigm for sociolinguistics and history, the study is carried out with secondary sources (linguistic databases, scholarly literature, demographic reports, and cultural literature). The findings of this research show that Kinyarwanda belongs to a very homogenous Rwanda-Rundi language group in the Bantu group of the Niger-Congo language family with many structural and phonetic correspondences and a high degree of intelligibility with other languages such as Kirundi or Rufumbira. The spread of Kinyarwanda across countries such as the Democratic Republic of the Congo, Uganda, Burundi and Tanzania suggests a unified language community in the region, rather than national language communities, and migration has helped Kinyarwanda to become a diaspora language. The paper refers to this situation of cultural identity through language as a theory of “Rwandafoni,” and compares it with other language groups in the world. The authors explain that people use language as a process which exists beyond the limits of national territory. Keywords: Kinyarwanda, Rwandafoni, Rwanda-Rundi Languages, African Sociolinguistics, Great Lakes Region, Transnational Identity, Diaspora Language, Bantu Languages. Title: Rwandafoni: The Transnational Linguistic Community of Kinyarwanda Across the African Great Lakes Region and the Global Diaspora Author: Professor Dr. Idi (Ras) Banamungu, Claude Bizumuremyi International Journal of Interdisciplinary Research and Innovations ISSN 2348-1218 (print), ISSN 2348-1226 (online) Vol. 14, Issue 2, April 2026 - June 2026 Page No: 64-69 Research Publish Journals Website: www.researchpublish.com Published Date: 29-May-2026 DOI: https://doi.org/10.5281/zenodo.20443710 Paper Download Link (Source) https://www.researchpublish.com/papers/rwandafoni-the-transnational-linguistic-community-of-kinyarwanda-across-the-african-great-lakes-region-and-the-global-diaspora
English contains a small but systematic class of DP-internal expressions whose categorical status is not well captured by the standard labels «determiner», «adjective», «quantifier», or «noun». Canonical members include «the other» and «the rest», and— crucially for the present theory – their closest structural neighbors in the «indefinite ordinal» family such as «the next», «the last», «the former», «the latter», and «the following». We argue that these expressions instantiate a single, typed configuration: a «Selector-Head Expression (SHE)», i.e., a D-licensed selector constituent of the form «[D the] + [Sel X]» that introduces a discourse-anchored selection variable and therefore requires «definite DP closure». The central constituency claim is that the determiner is not merely an external modifier of a nominal head: in selector readings, «the» is the minimal licenser that closes the selector, so that «[the + X]» forms the smallest closed unit, while numerals and NPs («two», «ten people») are specifications of that already closed selector constituent. Within the ULD-geometry, we treat TBH hosting as a spectral environment that forces entry-material into a structured three-region distribution (ordinate/subordinate/coordinate). SHEs are situated at the boundary between the subordinate quantifier/ordinal sub-shell and the coordinate linking regime: they are term internal operators that select complements (remainder) or indices (order), yet behave as «nominally deployable» DP constituents. We develop (i) a «Results box» containing a Proposition and Theorem that make the D-locking and remainder–index unification explicit, (ii) a «Distinguishing Predictions» box that separates the present account from nearby post-determiner or substantivized-adjective analyses, and (iii) a «classifier protocol» that turns the inventory into a reproducible empirical tool. A compact diagnostic table operationalizes five tests (standalone nominal use, numeral specification, partitive availability, selector-type stability, and D-lock strength) and produces a tiered classification (core/extended/peripheral). We extend the empirical discussion beyond English via a cross-linguistic survey of how languages realize the same closure requirement (articles, demonstratives, clitics, agreement, classifiers), and we outline testable neurolinguistic and computational consequences. In processing, D-locked selector closure predicts earlier parser commitments upon encountering «the + X» (reduced adjective-attachment ambiguity) and specific ERP/reading-time signatures. In computation, we propose minimal annotation guidelines and a detection pipeline for SHEs in dependency treebanks. The broader claim is that SHEs provide a concrete, reviewer-visible case where «articles function as closure operators beyond canonical noun heads», supporting a geometry-first&nbsp;methodology in which occupation patterns follow from the availability of empty localities (ghost-lattice) and closure constraints rather than stipulation.
Surzhyk, as a Ukrainian–Russian mixed subcode, has long been a subject of extensive research. Despite this, it remains an ambiguous phenomenon that elicits diverse assessments and attitudes. Within Ukrainian academic circles, a tradition has emerged of evaluating this communicative subcode through the prism of linguistic norms, predominantly characterising it as a negative phenomenon. However, in everyday communication, Surzhyk functions dynamically, reflecting the varied attitudes of its speakers. Significant shifts in its perception became particularly evident following the Russian Federation's full-scale invasion of Ukraine. Consequently, Russian-speaking individuals have increasingly adopted Surzhyk as a means of distancing themselves from the Russian language, which is widely perceived as the language of the aggressor. This article presents the results of a sociolinguistic study conducted in 2020–2021 and 2023–2024 among residents of the Odesa and Mykolaiv regions. The study examines the evolution of attitudes towards Surzhyk in the wake of the invasion, with particular attention given to the role of mass media discourse in shaping these perceptions. The analysis reveals a growing tendency towards a more favourable view of mixed speech; in the context of war, Surzhyk has come to symbolise solidarity and serves as an informal linguistic bridge for transitioning from Russian-dominant to Ukrainian-speaking communication.
Social media is the most popular platform for opinion expression. Sentiment analysis is the process of acquiring information about things, events and their characteristics out of people’s views, assessments and feelings. Opinion mining is an alternative term for sentiment analysis. In this paper, Enhancing Opinion Mining of Twitter Data with a Deep Convolutional Spiking Neural Network and Balancing Composite Motion Optimization (OMTD-DCSNN-BCMO) is proposed. Initially, the Twitter data are obtained from the Stanford Sentiment Treebank (SST-2) dataset. Then, the data is fed to the preprocessing. The pre-processing output is provided to extract the Radiomic features depending on the Residual Exemplars Local Binary Pattern (RELBP). The extracted output is provided to the feature selection for choosing ideal features using the Piranha foraging Optimization Algorithm. The selected features are provided to a Deep Convolutional Spiking Neural Network (DCSNN) for classifying Twitter data as negative, positive and neutral. Then, the DCSNN approach is optimized using Balancing Composite Motion Optimization (BCMO) for better performance. The efficacy of the proposed technique is examined using performance metrics and the method attains 23.32%, 26.07% and 28.51% higher accuracy and 21.92%, 15.03% and 19.15% lesser error rate are evaluated with existing approaches.
This article provides a comprehensive analysis of the stages of formation and development of terminology in the field of materials science from both historical and comparative perspectives. The study traces the evolution of materials science terms in close connection with the advancement of scientific thought, technological progress, and socio-cultural factors influencing scientific language. Particular attention is paid to the gradual transformation of terminology from early descriptive naming practices to highly standardized and systematized term structures used in modern scientific discourse. In addition, the article examines the periodization of term formation as interpreted by the Russian, English, and Uzbek schools of scientific terminology. Each tradition is analyzed in terms of its theoretical foundations, methodological approaches, and linguistic mechanisms of term creation. The Russian terminology school is discussed in relation to its emphasis on systematization and conceptual hierarchy, the English school in terms of internationalization and borrowing processes, and the Uzbek school with regard to term adaptation, nationalization, and linguistic modernization. The development of materials science terminology is explained step by step, highlighting key historical stages such as the initial borrowing of general scientific vocabulary, the emergence of specialized terms during industrial and technological revolutions, and the contemporary phase characterized by interdisciplinary integration and terminological globalization. Comparative analysis reveals both shared tendencies and language-specific features in term formation, including word-formation models, semantic shifts, and translation strategies. Overall, the article demonstrates that the formation and development of materials science terminology is a dynamic and continuous process, shaped by scientific innovation, linguistic norms, and cultural context. The findings contribute to a deeper understanding of terminological evolution and offer valuable insights for terminology studies, translation theory, and the standardization of scientific language across different linguistic traditions.
According to the traditional view, the metaphysician’s task is purely theoretical: to describe the fundamental structure of reality as it is, independent of human interests. By contrast, the work of Amie Thomasson—most notably her defense of "easy ontology" and her conception of metaphysics as conceptual engineering—presents a seemingly compelling case for the erosion of the Kantian distinction between theoretical and practical rationality. In this paper, we show how Thomasson’s deflationary approach to metaphysics effectively relocates metaphysical inquiry from the domain of pure theoretical rationality to the domain of practical reason. According to Thomasson, existence questions are answered by a combination of "application conditions" (conceptual/linguistic rules) and empirical facts. If the rules for using the term "table" are met by a certain arrangement of atoms, the existence of the table follows analytically. Consequently, the "hard" work of metaphysics is not a theoretical search for deep truths about the world, but rather a normative and practical (metalinguistic) negotiation over our conceptual framework. In Thomasson's view, many metaphysical disputes can, and should, be resolved by performing the right kind of conceptual engineering. This shift transforms the metaphysician into a conceptual engineer. We are going to explore three key ways Thomasson’s philosophy bridges the theoretical-practical divide in metaphysics: (1) metametaphysical deflationism and the reduction of ontological inquiry to the practical application of linguistic norms; (2) metaphysics as conceptual engineering, where the choice of a conceptual scheme is guided by practical goals, values, and the functions we need our language to perform; and (3) the normative dimension of metaphysics, where "theoretical" rationality in metaphysics is inextricably tied to "practical" reason—specifically, the normative task of deciding how we ought to categorize the world. Finally, by critically assessing Thomasson’s rejection of "heavyweight" metaphysics and her neopragmatist reconceptualisation of the field, we identify some significant weaknesses in her overall project and argue in favour of a more traditional conception of metaphysics as a genuinely theoretical enterprise.
Post-retrieval extinction has been proposed to modify fear memories by introducing safety information during the reconsolidation window. Imagery rescripting (ImRS) is a visual imagery-based intervention that alters the valence of emotional memories through UCS revaluation. The integration of ImRS and extinction targets both UCS revaluation and expectancy revaluation, providing a dual mechanism for fear reduction. This study explores the combined effects of imagery rescripting (ImRS) and extinction during the reconsolidation window to attenuate the return of generalised fear. Sixty-one healthy individuals underwent fear acquisition and generalisation (Day 1), followed by intervention (Day 2: Reactivation- ImRS and Extinction group or ImRS and Extinction group). On Day 3, return of generalised fear through reinstatement was tested using unsignalled UCS presentations. Skin conductance responses (SCR), expectancy ratings, and valence ratings were the outcome measures. The reactivation-ImRS + extinction group showed reduced reinstatement effects compared to the ImRS + extinction group, evidenced by attenuated SCR and increased CS+ valence ratings. The current study suggests that imagery rescripting and extinction conducted post-retrieval is more effective in reducing the return of conditioned fear. Therefore, integrating ImRS with post-retrieval extinction may offer a novel therapeutic pathway for reducing fear relapse.
This dataset contains 492 anonymized survey responses collected at seven urban coastal stops along the Antalya coastline, Türkiye. The data include perceived affective quality ratings based on the ISO 12913-2 soundscape framework, perceived dominance ratings for seven sound-source categories, spatial and temporal context variables, and overall satisfaction with the sound environment. No directly identifying personal information is included. The dataset supports the manuscript entitled “Modelling Urban Soundscape Satisfaction from Perceived Sound Sources: An Explainable Machine Learning Study along the Antalya Coastline.”
This article delves into the profound influence of linguistic culture on literary expression, exploring how writers utilize language to evoke meaning, capture authenticity, and reflect the complexities of human experience. Through an examination of various aspects, including language as a reflective medium, its role in evoking atmosphere and setting, and its exploration of cultural identity and diversity, this study highlights the significance of linguistic culture in shaping the richness and depth of literary works. Additionally, the article discusses how writers subvert linguistic norms and conventions to create innovative forms of expression and preserve oral traditions and heritage. Ultimately, it argues that the multifaceted role of linguistic culture enriches the tapestry of literature, offering readers a deeper understanding of the world and ourselves.
ABSTRACT Emotional episodes involve partially coordinated changes in experience and autonomic activity, and subjective–physiological coherence appears to vary with aging and physiological stress. Physiological accounts of emotional aging propose that age‐related changes in autonomic and interoceptive function reduce the informativeness of bodily signals, weakening the coupling between feelings and bodily responses. We test whether subjective–physiological coherence is lower in older vs. younger normoxic adults, lower in hypoxic vs. normoxic younger adults, and similar in hypoxic younger adults and older normoxic adults; in secondary mechanism‐oriented analyses, we examine whether heart‐rate variability and cardiac interoceptive accuracy are statistically related to group differences in coherence. We will test these questions in three groups ( N = 120): younger adults in normoxia, younger adults under acute hypoxia (FiO 2 ≈ 12%), and older adults in normoxia. Participants will view emotional images and provide trial‐wise ratings of affect and approach motivation while heart rate and skin conductance are recorded. Emotional coherence will be indexed primarily by within‐person correlations between subjective arousal ratings and autonomic reactivity, with approach motivation and valence examined as complementary subjective dimensions.
AI safety systems are increasingly deployed in high-stakes domains such as mental health, yet they are built on implicit assumptions about language that do not hold in practice. We argue that AI safety systems implicitly decide whose language counts as interpretable, and that risk expressed outside those norms is systematically overlooked. At the same time, AI systems can learn new language faster than their safety mechanisms can adapt to it. This creates a structural gap: while models may recognize emerging expressions, safety guardrails remain anchored to static linguistic norms. We introduce the concept of linguistic lag to describe the delay between the emergence of meaning within a community and its recognition by safety systems. Through examples and empirical observations, we show that semantically equivalent expressions of distress yield divergent safety outcomes when expressed in adolescent digital dialects. We argue that this is not a data coverage problem but a consequence of mismatched adaptation dynamics: language evolves through rapid, decentralized social diffusion, while safety systems evolve
Mnemonic Structure and Style in Irish Proverbs by Marcas MacCoinnigh represents a major contribution to modern paremiological and stylistic studies.This monograph is the fifth volume in the Proverbium Online Supplement Series and examines the structural and stylistic features that make Irish-language proverbs memorable.This study is long overdue and provides the first systematic linguistic and mnemonic analysis of Irish-language proverbs.As stated in the opening chapter, "proverbs and proverbial language occupy a distinctive and enduring place in the Irish linguistic tradition, with documented roots reaching back to Old Irish literature" (Mac Coinnigh 2025: 20).Despite this rich linguistic legacy, a comprehensive examination of the structural and stylistic features of Irish-language proverbs has never been undertaken.Drawing on both close qualitative reading and corpus-based description, the analysis identifies recurring patterns across the data and offers valuable insights for linguists, folklorists, paremiologists, and scholars of the Irish language.Irish-language material has remained peripheral to international paremiology, even as the field has broadened to include traditions as varied as Ancient Greek, Hebrew, Arabic, Spanish, Russian, and Yiddish.Mac Coinnigh engages critically with foundational research by international scholars in the field and demonstrates the ways in which Irish-language proverbial material both conforms to and diverges from cross-linguistic norms.
Language-model representations provide structured, high-dimensional annotations of naturalistic language stimuli and can serve as informative neural predictors during comprehension. We analyzed locked derived data from Brain Treebank, MEG-MASC, and Podcast ECoG with eight frozen language models, blocked encoding models, and matched temporal, nuisance, and representation-capacity controls. Positive held-out prediction and gains over low-level baselines were widespread in source-level summaries. Across Brain Treebank and Podcast ECoG, 67 of 432 evaluable rows met a controlled predictive-only criterion, and model-side feature ablations changed prediction scores in most evaluable source rows. Brain-derived, timing-linked, acoustic, and implanted-signal controls confirmed component-level sensitivity of the analysis pipeline. These findings show that language-model-derived quantities can annotate neural activity during natural speech and text comprehension. Participant-level matched-control advantages were localized rather than uniform, response-profile and feature-specificity contrasts bounded representational or computational interpretations, and complete co-indexed integrated interpretation will require future jointly indexed coverage. Together, the analyses identify language-model features as useful neural predictors and separate predictive usefulness from claims about shared neural organization or language-processing computations.
Inspired by general deviance theory, which proposes that individual traits predispose people to violate multiple types of norms or rules, we suggest that disregard for native language rules of speech may be positively correlated with disregard for other social norms such as truth-telling. Specifically, we hypothesize that given the opportunity to gain financially through untruthful statements, native-born speakers who speak incorrectly will be more likely to exploit it than their counterparts who speak correctly. We report the results of testing this hypothesis on Israeli students, using two double-stage experiments. The first stage, identical in both experiments, was intended to identify untruthful statements by applying the popular die-under-the-cup task. The second stage, aimed at identifying incorrect pronunciation, differed between the two experiments. In the first experiment, participants were asked to read aloud 10 sentences, whereas in the second experiment, conducted in writing, participants were presented with the same 10 sentences, this time phrased as questions with multiple-choice answers regarding the correct pronunciation. Our hypothesis that untruthful statements are positively correlated with incorrect speech was confirmed in both experiments, thereby providing empirical support for the notion that linguistic norm violations may extend beyond the domain of grammar and be linked to broader patterns of social norm disregard, such as truth-telling.
Digital communication has profoundly transformed the nature of textuality in both English and Azerbaijani contexts, affecting language structure, stylistic conventions, and the interactional norms of written discourse. The evolution of online platforms—including social media, messaging applications, and digital forums—has encouraged brevity, immediacy, and multimodality in textual expression. English, as a globally dominant language in digital spaces, exhibits widespread use of abbreviations, emojis, hashtags, and memes, reflecting both linguistic economy and the need to convey affective meaning rapidly. In contrast, Azerbaijani, while increasingly influenced by these global digital trends, maintains certain structural and grammatical particularities, including agglutinative morphology and context-dependent sentence constructions, which interact uniquely with digital shorthand and visual markers of communication. Comparatively, English digital texts often prioritize directness and global comprehensibility, leveraging standardized syntax and widespread conventions that facilitate cross-cultural interaction. Azerbaijani digital communication, while adapting similar strategies, demonstrates a more hybrid textuality: users negotiate between traditional linguistic norms and the pressures of digital brevity, producing hybrid forms that mix Latin-based abbreviations, code-switching, and local syntactic patterns. These patterns reveal not only the flexibility of the Azerbaijani language in accommodating technological mediation but also highlight the interplay between digital literacy, cultural identity, and language preservation.
Digital mental health is undergoing rapid transformation through the deployment of artificial intelligence (AI) systems including conversational agents, risk-prediction tools, and clinical decision-support platforms. Despite growing enthusiasm, the field has yet to converge on an operational standard for what constitutes honest, trustworthy AI in psychiatric and psychological contexts-one that explicitly accounts for the limits, blind spots, and populations for whom deployed systems have not been validated. This Perspective draws on Richard Feynman's foundational principle of scientific integrity-articulated in his 1,974 Caltech commencement address, "Cargo Cult Science"-and Carl Sagan's dimensional metaphors to develop a practical, disclosure-oriented ethical framework for AI in digital mental health. The neurodiversity paradigm provides a critical stress test: growing empirical evidence indicates that AI tools trained on narrow behavioural and linguistic norms carry substantial risk of misrepresenting neurodivergent users, with the potential to silently re-encode stigma, misclassification, and inequity into clinical systems. Drawing on predictive processing theory and recent empirical evidence of AI bias against neurodivergent populations, this work proposes five criteria for the Feynman Honesty Standard: scope clarity, population-limit disclosure, uncertainty communication, role integrity, and participatory co-development. The analysis concludes with a forward-looking research agenda for clinicians, developers, and policymakers, arguing that the field will be judged not only by algorithmic performance but by its candour about what these systems know, miss, and cannot do.
Abstract Vision–language models (VLMs) increasingly evaluate visual content, yet their behaviour across cultural traditions remains poorly characterised. We show that two open-weight VLMs, Qwen3-VL-8B and Llama-3.2-11B-Vision, assign systematically lower scores to East Asian art than to Western art, with Cohen’s d=−0.46 and −0.36 (p < 10−16). Internal probing reveals this disparity is accompanied by higher encoding cost under the primary reference sentence, elevated perplexity, reduced token confidence, and increased hidden-state norms. Across the two tested models, fifteen of eighteen signals shift in the same direction, indicating a consistent cross-model pattern; Llama cross-attention entropy shows a large effect (d=+1.30, p < 10−178). The gap persists across decoding temperatures (T ∈ {0, 0.5, 1}; all p < 10−7), suggesting that it is not driven solely by stochastic decoding. Hidden-state clustering places East Asian art in broadly overlapping representational regions relative to Western art, and keyword-based response analysis shows frequent Chinese lexical attribution for Korean artworks (88% Qwen, 65% Llama). In Qwen, matched-complexity analysis shows 87% of the score gap persists after controlling for image spectral properties. These results show that internal probing can help detect cross-cultural processing differences in the tested VLMs.
English is often described as the current universal language—or lingua franca—of science, much as Latin was during the Early Modern period (late 15th–18th centuries); however, this claim is open to question. Two fundamental differences can be identified. First, Latin was adopted as the scholarly language because the accumulated erudite knowledge of the time had been expressed in that language. For this reason, the renowned Italian physicist Galileo Galilei1 (1564–1642) published his first major work on astronomy in Latin, Sidereus Nuncius (1610). Similarly, the eminent British scientist Isaac Newton (1643–1727) published his seminal work in Latin, Philosophiæ Naturalis Principia Mathematica (1687). English, by contrast, became dominant for very different reasons linked to the power dynamics of dominant nations following the Second World War. Prior to the dominance of English, particularly during the nineteenth and early twentieth centuries, other languages were also used for scientific communication and enjoyed comparable acceptance within the scientific community. For example, Louis Pasteur (1822–1895) published his work in French. Alessandro Volta (1745–1827) published in Italian and French, and Enrico Fermi (1901–1954) likewise published in Italian before emigrating to the United States. Similarly, three major milestones in physics were published in German: Max Planck′s (1858–1947) postulate for quantum theory (1900), Albert Einstein′s (1879–1955) two theories of relativity (1905, 1915), and Erwin Schrödinger′s (1887–1961) fundamental equation of quantum mechanics (1935). Likewise, Santiago Ramón y Cajal (1852–1934), Nobel Prize laureate in Medicine in 1906, published his groundbreaking discoveries on neurons in Spanish. The second difference is that, when Latin served as the universal language of science, it was neither a vernacular language nor one spoken in everyday life by the general population. By the Middle Ages, Latin had evolved into the various Romance languages, including Spanish, Portuguese, French, Italian, and Romanian. Precisely because Latin was already a dead and immutable language, it was considered suitable for the universal communication of scholarly knowledge. In this historical period, Latin can therefore be understood as an authentic lingua franca, which, by definition, is not a native language of any of its users (Carli and Calaresu 2017). Thus, until the eighteenth century, anyone wishing to approach science, understand it, and communicate about it had to learn Latin. English, by contrast, should not be considered a lingua franca of science, as it is currently the native language of a significant number of scientists, granting them an evident advantage over those for whom it is not. In fact, a recent study has found that the use of English as the common language of science represents a major impediment to maximizing the contribution of non-native English speakers to scientific research, particularly in the early stages of their careers (Amano et al. 2023). This communicative disadvantage can be interpreted as a case of what Miranda Fricker terms epistemic injustice (Fricker 2007), a concept that has been examined in the context of organizations (Wilmot 2024), marginalized or underprivileged social groups (McConkey 2004), institutions of science (Grasswick 2017; Vučković and Sikimić 2023), and multilingual classrooms (González-Howard et al. 2023; Kerfoot and Bello-Nonjengele 2023). In what follows, the latter two areas of epistemic injustice within the context of science education are briefly discussed. In school, epistemic injustice emerges when monoglossic policies compel learners to acquire and demonstrate knowledge in a single language that is not familiar or native to all students. For example, Deng et al. (2022) found that in English-medium chemistry classes with students from diverse linguistic backgrounds, those with lower English proficiency demonstrated poorer scientific argumentation in their written work compared to their peers with higher proficiency. Similarly, Liu et al. (2021) reported that non-native English-speaking university students in Arts and Sciences participated less in collaborative learning and co-curricular activities than their native-speaking peers and showed smaller gains in objectively assessed critical thinking skills. These findings illustrate the tangible consequences of linguistic inequities in educational settings and highlight the need for pedagogical approaches that recognize and value students′ full linguistic repertoires. As an alternative to address this linguistic problem, Kerfoot and Bello-Nonjengele (2023) propose adopting multilingualism as a point of departure, legitimizing all languages within students′ repertoires in order to reduce epistemic hierarchies. This perspective aligns with translanguaging pedagogies, which advocate allowing students to express what they know and are able to do across their full communicative and semiotic repertoire (Fine and Furtak 2025). In this regard, Charamba (2021) found that the implementation of translanguaging approaches was associated with increased student participation in scientific practices, improved conceptual understanding of scientific content, enhanced confidence and epistemic agency—particularly among students previously marginalized by monolingual policies—and a reduction in linguistic hierarchies, thereby promoting greater epistemic equity in the science classroom. Similarly, González-Howard et al. (2025) concluded that intentionally and reflectively integrating multiple languages into science teacher education is a necessary condition for advancing more just, inclusive, and transformative science education, both in teacher preparation and in future practice with multilingual learners. So, if research in science education has established strategies to address epistemic injustice arising from multilingualism in the classroom, could these approaches serve as inspiration for addressing similar forms of epistemic injustice within the field of science education research itself? Without attempting to directly compare school and academic contexts in terms of linguistic challenges, and drawing on previous studies (Amano et al. 2023; Carli and Calaresu 2017), it can be argued that science education researchers whose native language is not English generally have to make a considerably greater effort than their native English-speaking peers to internationalize their work. Beyond the inherent difficulty of scientific communication (Petzold et al. 2025), this process requires mastering field-specific terminology and expressions in English, as well as identifying their equivalents in one′s own language. It should be emphasized that, unlike the exact and experimental sciences, science education is a social science; therefore, argumentation and narrative play a particularly important role in research. As a direct consequence of these challenges, translating research into English often compels authors to relinquish certain nuances, expressions, and stylistic features in order to make their texts comprehensible to an Anglophone audience. This frequently raises the question of how many research studies in science education may have failed to achieve the international recognition they deserved simply because they were not written in English. Indeed, using a single language for the communication of scientific research entails an inherent loss of the diversity of ideas that emerge within the plurality of languages (Vučković and Sikimić 2023). Therefore, linguistic communication barriers constitute a clear factor of epistemic injustice within the field of science education research. The linguistic obstacles are particularly pronounced at international forums for science education research, where presentations and discussions are conducted exclusively in English. Native English-speaking researchers are exempt from this challenge, as they produce research and deliver lectures in their mother tongue, an act predicated on the assumption of universal comprehension. Notably, this dynamic persists even when such conferences are hosted in non-Anglophone nations, with English remaining the default language of presentation, often supported only by ad-hoc simultaneous interpretation for attendees. In some instances, even this basic accommodation is absent, leaving non-native speakers at a profound disadvantage. However, to the best of my knowledge, researchers whose native language is not English are never allowed to deliver a keynote in their own language at international science education conferences organized by or held in Anglophone countries. From an equity perspective, would it not be fair for non-English-speaking researchers to present in their native language, with simultaneous interpretation provided for those who do not understand it? Several years ago, Siu Ling Wong and Derek Hodson published an interesting article in Science Education entitled “From the horse′s mouth: What scientists say about scientific investigation and scientific knowledge” (Wong and Hodson 2009). When I first read the title, I did not understand what was meant by the expression “from the horse′s mouth.” Its literal translation into Spanish made no sense to me. I therefore had to investigate its meaning, discovering that it is a British idiom referring to obtaining information directly from the most reliable source. I would never have imagined that a horse could be considered a source of reliability. In Spanish, we also have figurative expressions with the same meaning, such as “saber algo de buena tinta” (“to know something from a good ink”) or “saber algo de buena mano” (“to know something from a good hand”). However, if a Spanish researcher were to use literal translations of these expressions in the title of an English-language paper, it is very likely that the expressions would be rejected or that the author would be advised to replace them with something more “natural” in English (e.g., “to know something from a reliable source”). Allowing Anglophone authors to use idiomatic expressions from their own language in scientific publications thus constitutes a clear linguistic advantage for them and an epistemic disadvantage for others. Many expressions and nuances became obstacles for non-English-speaking science education researchers when we first began writing in English. For example, in the context of Ibero-American countries we speak of “didáctica de las ciencias,” which could be literally translated to English as “didactics of science,” to refer to the field of knowledge and research devoted to improving the teaching and learning of science. In English, however, the expressions “science education” or “science teaching” are typically used. Yet in Spanish, these latter terms do not have the same meaning as “didactics of science.” “Science education” or “science teaching” refer to the enactment of the science curriculum in the classroom, regardless of the methods, approaches, and resources used, whereas “didactics of science” denotes the academic field that studies those aspects. In their discussion of language and science, Carli and Calaresu (2017) have already addressed this by examining how certain scientific fields are named and/or conceptualized depending on whether they are situated in an Anglophone linguistic context or in other contexts, such as continental Europe. The dominance of English also has implications for the terminology used in science education in non-English-speaking contexts, as insufficient attention is paid to the careful adaptation of certain terms to these contexts. For example, some concepts recently incorporated into the Ibero-American science education context include “epistemic agency” (Stroupe 2014) and “student agency” (Arnold and Clarke 2014), which have been translated literally as “agencia epistémica” and “agencia estudiantil” (or “agencia del estudiante”). In my view, however, these translations are inappropriate. In English, “agency” has three main meanings: (1) an organization or business that provides a service, (2) action or intervention that produces an effect, and (3) the capacity to make decisions and act independently. In Spanish, however, according to the Royal Spanish Academy, the term “agency” (“agencia” in Spanish) corresponds only to the first of these meanings. Consequently, for a Spanish speaker, “student agency” may simply suggest that the student belongs to some kind of organization, rather than conveying the intended sense of autonomy, decision-making, and capacity for action. This semantic mismatch becomes particularly problematic when these terms are examined in their original theoretical context. In English-language science education literature, “epistemic agency” refers to the capacity to shape, produce, evaluate, and actively use knowledge rather than passively receive it (Zhang et al. 2022), while “student agency” refers to students′ proactive and transformative participation in classroom activities (Mameli et al. 2023). Both uses clearly correspond to meanings (2) and (3) of the English term “agency.” Accordingly, in Spanish context “epistemic agency” should have been translated as “capacidad epistémica” (i.e., “epistemic ability”) or “acción epistémica” (i.e., “epistemic action”) depending on the context in which it is used, and “student agency” as something akin to “capacidad del estudiante para la acción transformadora” (i.e., “students′ ability to engage in transformative action”). Yet this has not occurred, and we already find works in Spanish that have adopted these literal translations (e.g., Costa Ramos et al. 2021). A reverse linguistic imposition would surely not have taken place. All of the above reflects some of the linguistic experiences of a native Spanish-speaking researcher; however, in light of the findings reported by Amano et al. (2023), it is likely that other researchers whose first language is not English face similar challenges in their scientific communication practices. Moreover, translation issues such as those discussed above should not be regarded as merely anecdotal, but rather as shaped by implicit evaluative norms in science education research that privilege Anglophone conceptualizations and linguistic conventions—norms that operate not only in the writing of research articles but also in their review and editorial decision-making processes. As such, these issues point to a broader pattern in the field, whereby English-based linguistic norms influence what is considered legitimate, precise, or conceptually acceptable knowledge. In Fricker (2007) terms, this situation constitutes a form of hermeneutical injustice, as dominant Anglophone frameworks constrain the interpretive resources available to non-English-speaking researchers, limiting their capacity to fully articulate their perspectives and have their knowledge adequately recognized. None of that is intended to suggest that non-native English speakers should stop using or publishing in English, which, of course, we will continue to do with conviction. Rather, it is an invitation for Anglophone researchers to also make the effort to disseminate their findings in languages other than their own, and for the international science education community to ensure that all researchers have the opportunity to publish or present their work in the language they master best. As Kersting et al. (2025) argue, establishing multilingualism and multimodality as structuring principles of science education is key to ensuring equitable, socially relevant, and conceptually deep learning, especially in linguistically and culturally diverse contexts. There are currently sufficient means and resources to make this feasible in science education research. Vučković and Sikimić (2023) frame all of this as an epistemic equity challenge, implying that researchers disadvantaged by their linguistic abilities (Amano et al. 2023) should receive support from the broader scientific community. This support could take the form of mitigating agents—individuals or organizations that act as a communication bridge between individual researchers and the scientific community (Vučković and Sikimić 2023). Given all these considerations, and in order to move toward a more epistemically just international research landscape, I contend that the key lies in ceasing to privilege English and instead seeking to reconcile the use of multiple languages for communicating debates and research findings in science education. Following Carli and Calaresu (2017), an ideal scenario for achieving linguistic equity in a multilingual scientific communication context would be one in which researchers—including in their roles as reviewers and editors of academic publications—possess sufficient skills to understand work published in languages other than their own (at least those closely related or widely spoken),2 while also having the option to disseminate their research in their native or predominant language. If publication editors then consider it appropriate or necessary to translate a work into English, they should do so without imposing any additional cost or workload on the researchers. Today, sufficient resources exist to overcome many lexical-semantic barriers across languages: artificial intelligence enables translations that largely preserve characteristic nuances of each language, and simultaneous interpretation has long facilitated communication and debate at international scientific events. In light of this, does it still make sense to continue prioritizing English as the primary language for maximizing the international reach of science education research? In the Ibero-American context, for example, this issue has largely been overcome, as most journals allow publication at least in Spanish, Portuguese, and English. Moreover, many of these journals are indexed in leading bibliometric databases such as SCOPUS and/or Web of Science—for example, Revista Eureka sobre Enseñanza y Divulgación de las Ciencias (Spain), Revista Brasileira de Ensino de Física (Brazil), and Educación Química (Mexico). The question is: when will Anglophone journals take this step? Similarly, science education research conferences in the Ibero-American context commonly allow presentations in both Spanish and Portuguese, while opening or closing lectures may also be delivered in English if it is the speaker′s first language. In contrast, it is striking that an international event such as the successive editions of the ESERA (European Science Education Research Association) conference permits presentations only in English, excluding all other official languages of the European Union. This decision likely reflects political rather than budgetary considerations. Moreover, a significant proportion of attendees are native speakers of these other non-English languages. For instance, Spanish alone is spoken officially in 21 countries and by over 600 million people worldwide, including nearly 60 million in the United States. This could also prompt the organizers of NARST (National Association for Research in Science Teaching) conferences to in both Spanish and English. I not the of English in that but rather issues of equity and epistemic at a educational institutions should be to these When we have a language, which is not a lingua franca for a of the scientific community does not have to time and resources to learn the languages and that time and resources which the will use to learn and to use the language can be used by native speakers for their research a international language, the of is therefore, but is simply from the community to a of the community. It is that, at a time when science education to such as epistemic and implications (Grasswick 2017; Fricker et al. when language is as a factor and Bello-Nonjengele 2023; Vučković and Sikimić 2023), this has still not been taken in science education research. the issue to a of key of science the to be is that a and debate on this issue can within the international science education research including researchers, and organizers of scientific It is that some of the ideas discussed may to advancing this writing original writing review and intelligence have been used with the of improving the writing of the All is the work of the
This article examines the significance of observing linguistic norms in translation from Russian into Uzbek. It analyzes the historical background of the interaction between the two languages, the typological differences between synthetic Russian and agglutinative Uzbek, as well as the main types of norms: lexical, grammatical, stylistic, and orthoepic. Special attention is paid to the problems of conveying non-equivalent vocabulary, phraseological units, realia, and metaphors. Specific examples illustrate the difficulties encountered in translation and ways to overcome them to achieve translation adequacy and equivalence.
The subject of this study are the neologisms within the lexical-semantic field INTERNET. The object of the research are the ways of their adaptation in modern Russian and their features. The article discusses issues related to the change of the attitude of internet user toward the need to comply with linguistic norms, which causes the emergence of orthographic, phonetic, morphological and word-formation variants of loanwords. The aim of this article is to classify the adaptation features of English-language lexemes, namely, to identify their variants belonging to different aspects of language. To achieve this purpose, the following objectives were set: to identify borrowed nouns within lexical-semantic field INTERNET, to study their functioning in various contexts, and to classify the features of lexeme adaptation based on this analysis. Modern explanatory dictionaries, as well as search data from Google and Yandex, were used as sources of linguistic information on the adaptation of loanwords. In order to identify the functional characteristics of these units in modern Russian, the descriptive method was used in this article. The units were also characterised from a stylistic perspective, the features associated with the use of different alphabets – Cyrillic and Latin – were also examined. The scientific novelty of this study lie the systematic analysis of loanword adaptation using units from the lexical-semantic field INTERNET. This analysis identified key areas of variability at different language levels, and concluded not only that the ongoing adaptive changes are systematic but also that the scope of variability is expanding. This is largely due to the rapid growth in the number of borrowings, resulting in the emergence of new graphic and phonetic variants. It is also noted that, although units from this lexical-semantic field are gradually being reflected in explanatory dictionaries of the modern Russian language, the fixed use of words in dictionary definitions does not always lead to a reduction in variability.
English as a Lingua Franca (ELF) has become the primary medium of international communication, yet language assessment practices continue to privilege native-speaker norms that misalign with multilingual communication realities. This study addresses the critical need to redefine oral assessment frameworks for ELF contexts, particularly in Japanese universities where English increasingly serves as a cross linguistic tool among diverse multilingual speakers rather than as an endpoint modeled on native-speaker competence. Using a two-phase qualitative design, this study analyzed seven naturally occurring dialogues (10,030 words, ~ 2.35 h) from the English as a Lingua Franca in Japan (ELFJ) Corpus involving Japanese and non-Japanese English speakers. Conversation Analysis and Interactional Sociolinguistics were employed to identify successful ELF communication strategies. Document analysis of Cambridge English assessment frameworks was conducted to examine how existing competency descriptors could be reconceptualized for ELF-aligned (context-valid) assessment contexts. Nine distinct ELF communication strategies emerged as essential for successful multilingual interaction: negotiation of meaning, self-initiated communication strategies, simplification, repair, lexical innovation, code-switching, accommodation, turn-taking management, and collaborative construction. Analysis of existing assessment frameworks revealed three critical inadequacies: native-speaker norm orientation, limited recognition of strategic communication, and absence of multilingual resource recognition. Based on these findings, a reconceptualized framework with four ELF-informed criteria is proposed for future validation: Intelligibility and Clarity, Strategic Communication, Interactional Effectiveness, and Multilingual Resourcefulness. This study suggests that ELF-aligned (context-valid) oral assessment requires fundamental reconceptualization beyond superficial modifications to existing frameworks. The proposed ELF-informed assessment criteria prioritize communicative effectiveness and strategic flexibility over adherence to native-speaker norms, potentially offering a more valid and equitable approach to evaluating multilingual speakers’ English proficiency pending empirical validation. These findings have significant implications for language testing policy and practice in increasingly multilingual educational contexts across Asia and globally.
The article presents a critical analysis of the curriculum for the discipline “Foreign Oriental Language (Chinese Language)” designed for the bachelor’s degree program No. 60230100 — Philology and Language Teaching (Oriental Languages), implemented at the Tashkent State University of Oriental Studies (2024). The relevance of the study is determined by the growing need to rethink the content of Chinese language education in the context of expanding intercultural contacts and the increasing role of the language as a means of professional, academic, and intercultural interaction. The author aims to determine the extent to which the principles of the cognitive- discursive approach and components of intercultural communication are implemented in the analyzed curriculum, as well as to identify priority areas for its further methodological and content-related improvement. The study employs methods of analysis of normative and methodological documentation, content analysis of instructional materials, and comparison of educational objectives with the declared learning outcomes. The curriculum is generally oriented toward the step-by-step formation of language competencies at levels A1–C1 and includes lexical-grammatical, analytical, and research elements that ensure the development of the main types of speech activity. At the same time, the article demonstrates that these elements are predominantly aimed at developing formal linguistic skills and do not sufficiently contribute to the development of discursive competence, cognitive strategies of meaning comprehension, and cognitive-analytical and interpretative strategies of working with texts. According to the author, the intercultural component of the curriculum is presented fragmentarily and is not formulated as an independent learning objective: there is no systematic work with culturally marked discourses, and skills of comparing communicative norms, reflection, and adaptation to intercultural differences are not developed. In conclusion, the necessity of transitioning from a predominantly language- centered model of instruction to a cognitive-discursive and intercultural paradigm is substantiated, aimed at forming foreign-language consciousness and preparing students for effective communication in a multilingual and multicultural environment
The phenomenon of lexical borrowing plays a central role in the evolution of languages, reflecting cultural interaction, globalization, and social development. Both modern English and Uzbek have adopted a significant number of foreign lexical items, many of which undergo notable semantic changes after being integrated into the recipient language. This study investigates the patterns and mechanisms of semantic shifts in English and Uzbek loanwords, with particular emphasis on processes such as semantic narrowing, semantic broadening, amelioration, and pejoration. Examples drawn from literary works, media sources, and academic texts demonstrate how borrowed words adapt to local linguistic and cultural contexts. The analysis reveals that English loanwords generally preserve their original meanings due to global standardization, whereas Uzbek loanwords often experience substantial semantic modification influenced by local usage, social factors, and cultural norms. This research provides insight into the dynamic nature of lexical borrowing and its role in language development.
Introduction. The relevance of this study stems from linguistic interest in the cross-cultural description of conceptual value formations, including those of a universal nature, such as moral and ethical values. The novelty of this study lies in its identification of the culturally specific content of the value of "Loyalty/Devotion" in Arabic and English cultures. Methodology and sources. The study utilized corpus linguistics and interpretive discourse analysis. The empirical basis for the study was the Leeds Corpus of Contemporary Arabic; the English corpora COCA (Corpus of Contemporary American English); The British National Corpus (BNC1994; BNC2014) includes texts of various genres. Corpus methods were used to identify the frequency of lexical units representing the concept, describe their collocation profile, analyze contextual meanings, and define a cognitive-propositional model of value. Interpreting the results in the context of Max Weber's theory of social action revealed value implications reflecting behavior patterns characteristic of different linguistic cultures. Results and discussion. It was established that in English linguistic culture, devotion/loyalty is correlated with Weber's goal-oriented rationality, is pragmatically determined, and is oriented toward an external demonstration of commitment, expressed through specific actions and compliance with institutional rules, norms, and laws. In the Arabic linguistic culture, which is oriented toward value-based and rational action, loyalty/devotion is based on intrinsic motivation and a sense of duty, which determine the value orientations of members of the discourse community. Conclusion. The proposed methodology can be used in the analysis of other moral and ethical values in terms of their content and discursive-textual processing, as well as in the cross-cultural study of values to identify their universal and specific features.
Writing is a technical talent that calls for proficiency with both linguistic and organizational elements. The primary objective of this study is to examine the most common mistakes in argumentative essays written by ninth-grade students on an issue of "Advantages and Disadvantages of School Uniform." A total of ten essays were gathered in a controlled classroom setting and evaluated using an error analysis approach. The findings reveal that punctuation problems are most common (31.1%), followed closely by cohesion defects (25.9%) and distantly by grammar (21.8%) and lexical (21.2%) errors. The results indicate that students' most struggle comes with text arrangement and writing mechanics in comparison to what they do with grammar proficiency itself. The study highlights the critical importance of targeted pedagogical interventions to improve students' overall writing competence. Introduction Writing is typically recognized as one of the most demanding skills in English as a Foreign Language (EFL) teaching. It requires learners to incorporate grammatical knowledge, vocabulary, punctuation, and logical sequencing into a coherent written work. Writing an argumentative essay faces an additional challenges for secondary school students, particularly those in the ninth grade, as it involves opinion expression, idea comparison, and maintenance of balanced structure. The topic "Advantages and Disadvantages of School Uniform" is commonly used in the classroom practice since it facilitates student's expression of both positive and negative perspectives. However, despite their familiarity with the topic, many students struggle to generate accurate and well-organized essays. Error analysis offers valuable insights into learners' language development and identifies areas that require instructional support. Therefore, this study aims to address the following research question through investigating the most frequent mistakes made by students in their writing: What are the primary linguistic obstacles that ninth-grade EFL students confront while composing an argumentative essays? Literature Review Error analysis has historically been acknowledged as a crucial approach in second language acquisition. According to Corder (1967), errors are natural and essential component of learning process since they reflect learners' developing competence. Similarly, Brown (2007) highlights that mistakes provide crucial details about how language norms are adopted by students. Hyland (2003) emphasizes that mastery of discourse structure genre rules is just as important to good writing as grammatical correctness. In argumentative writing, students must use cohesive to guide the reader and properly arrange concepts to guide the reader. Ellis (1997) states that learners often undergo struggles in multiple areas simultaneously, including grammar, vocabulary and coherence. Prior studies have demonstrated a frequent trouble with cohesion and punctuation, which significantly affect the clarity and readability of the writing. Methodology 3.1 Reaserch Design This investigation implements both qualitative and quantitative error analysis techniques to examine students' writing. 3.2Participants Ten ninth-grade students learning English as a foreign language participated. Their competence level ranged from lower-intermediate to intermediate. To guarantee anonymity, students were labeled S1-S10. 3.3 Task Design Students completed a precise argumentative writing in a classroom setting. Topic: Advantages and Disadvantages of School Uniform Word limit: 120–180 words Time limit: 20–25 minutes Conditions: No dictionaries or translation tools were allowed This method verified for the accurate data representation of learners' autonomous writing. 3.4 Data Collection and Analysis A total of ten essays were collected and manually analyzed. Errors are divided into four categories. Grammar: subject-verb agreement, article usage, sentence structure Lexis: vocabulary choice, repetition, incorrect word forms Cohesion: linking devices, logical flow, organization Punctuation: commas, full stops, capitalization Calculations were made for combination of frequency counts and percentage distributions. Results 4.1 Quantitative Findings The analysis identified a total of 193 errors across the students' essays. Punctuation: 60 errors (31.1%) Cohesion: 50 errors (25.9%) Grammar: 42 errors (21.8%) Lexical: 41 errors (21.2%) 4.2 Summary of Findings The findings indicate that punctuation is the most troublesome area, followed by cohesion, while grammar and lexical errors occur at similar but slightly lower rates. Discussion The results obtained suggest that punctuation errors represent the most significant struggle for students. Frequent errors in capitalization, commas, and full stops indicate that learners lack a clear understanding of basic writing strategies. This weakens overall readability and results in ambiguous sentence boundaries. The second major issue lies on cohesion, stresses students' challenges in organizing ideas logically. Many essays fail in clear paragraph structure and appropriate linking words. Instead of presenting a string of paragraphs as well as well-developed arguments, a series of disconnected sentences are produced. This affirms Hyland's (2003) view that writing competence involves more than grammar; it requires control over discourse structure. Grammar errors, accounting for 21.8%, remain a notable issue. Subject-verb agreement, incorrect sentence construction, and misuse of articles are common issues. These errors are characteristics of typical developmental patterns that Brown (2007) described. Similarly, lexical errors point to a narrow vocabulary. According to Ellis (1997), who highlights the significance of vocabulary in efficient communication, students often rely on simple and repetitive words, which limits their ability to express complex ideas. Overall, the findings imply that mechanics (punctuation) and structure (cohesion) rather than grammar alone are the main causes of students' writing challenges. Conclusion This study investigated the writing errors of secondary school EFL students and identified four major categories of difficulties. The findings reveal that punctuation (31.1%) and cohesion (25.9%) are the most problematic areas, followed by grammar (21.8%) and lexical errors (21.2%). These outcomes imply that a comprehensive approach that goes beyond teaching grammar is needed to improve students' writing. Greater emphasis should be placed on teaching punctuation rules, developing cohesive writing skills and expanding vocabulary. Consequently, this pattern may have broader impact on students' academic performance since clarity, coherence, and task achievement directly demand for strong punctuation and cohesion. To address these issues, it is recommended that teachers: 1) provide explicit instruction on punctuation and sentence boundaries; organize activities that develop cohesion and logical organization; encourage the use of diverse vocabulary offer regular, detailed feedback on students' writing These techniques can significantly enhance students' capacity to produce precise, coherent and well-structured argumentative essays.
The article is devoted to the study of the category of expressiveness and the peculiarities of its realization in the texts of articles within contemporary Ukrainian scientific discourse. The main theses focus on substantiating the regularities of expressing expressiveness and the dominance of syntactic expressive means in scientific discourse texts, as well as on revealing the specificity of the functioning of general linguistic and specific means of expressiveness in texts of a linguistic profile in light of enhancing their influence on the interlocutor. The main research method is descriptive, along with contextual analysis, which made it possible to identify the specificity of expressing expressiveness in scientific discourse articles. In addition, the hypothetico-deductive method and the method of comparative analysis were applied, allowing for the identification of common and specific features in the use of expressive means in scientific linguistic texts. The initial hypothesis of the study is the assumption that in modern scientific humanitarian discourse, the role of expressive influence is increasing. This is обусловлено social, political, economic, and subjective communicative-pragmatic factors and is manifested in the obligatory use of linguistic expressive means reflecting stereotypes of speech communication in the field of science. At the same time, syntactic means of expressing expressiveness are dominant in scientific discourse texts, since the use of lexical means is limited by stylistic norms. The categorical features of scientific discourse – intertextuality, dialogicity, the specificity of pragmatics, and the participants of communication – determine the genre differentiation of communication models. At the same time, the genre of the scientific article should be considered central in scientific discourse, since it not only contains the manifestation of all categorical features of scientific discourse but also represents an operative genre that fully meets the requirements of modern scientific cognition and communication. The analysis of the linguistic expression of scientific articles is relevant for determining the peculiarities of linguistic means of the scientific style and the specificity of scientific discourse.
This article examines the specific features of the conceptual architecture of the PSYCHOTRAUMA concept in contemporary English-language media-cultural discourse. The relevance of the study stems from the global shift in the mass film industry from a purely entertainment-oriented paradigm to one that serves a therapeutic function, where “quality television” becomes a safe space for the verbalization and destigmatization of acute mental crises. Based on the script of the first season of the series “The Bear,” a study of the concept’s nominative field (243 units) was conducted, enabling the reconstruction of the linguistic experience of trauma in the extreme conditions of a professional environment. The methodological framework of the study integrates three stages of interdiscursive analysis, ranging from static lexicographic norms and scientific terminology to dynamic script-based verbalization. The scenario-based structure encompasses seven lexical-thematic groups forming three macrofields: internal mental and psychosomatic states, professional abuse and kitchen chaos, as well as family dysfunction and defense mechanisms. A characteristic feature of the architectonics is the heterogeneity of the groups, in which culinary jargon (e.g., Yes Chef, fire, behind) serves to suppress pain and exercise compulsive control. Particular attention was paid to the quantitative parameters of the nominative field, which demonstrate the dominance of affective-obscene units. It has been established that the maximum frequency level is formed by the nominative units fucked up, motherfucker, fuckin' idiot (f=64), which, in the context of chronic stress, lose their invective function, transforming into verbal markers for immediate emotional release. In contrast, the low-frequency level is represented by unique somatic metaphors of destruction (floor collapsing, delicate fuckin' ecosystem) and descriptions of extreme physiological states (trouble breathing, throwing up every day before work), which capture individual aspects of pain. A synthesis of the results indicates a trajectory of the concept’s sequential transformation – from an abstract definition (emotional shock) and clinical objectification (PTSD, prolonged grief) to a subjective affective reality. It has been demonstrated that media-cultural discourse is the most optimal medium for reflecting the dynamics of psychological trauma, where everyday, often aggressive linguistic expression reflects an individual’s behavior in crises.
The Canvas Model's equality processor operates in two complementary feed-modes. Feed-backwards (Steering) is anticipatory and introverted: the processor compares the current state against stored experience, projects forward, and adjusts before error materializes. Feed-forward (Driving) is reactive and extroverted: the processor engages directly with the input, acting on what is present without consulting stored experience. The first four papers in this series explored the Feed-backwards mode—baseline subtraction as regularization, the S-invariant attractor for convergence, energy separation for modular architectures, and meta-time continuous training dynamics. All four operate in the anticipatory, introverted mode: learning from accumulated experience, correcting toward equilibrium. This paper explores the Feed-forward mode. What happens when the processor does not consult the past? It acts. It drives. What this paper provides: · A formal definition of the two feed-modes. Steering (Feed-backwards): d\mathcal{E}/d\tau = -\kappa \nabla_{\mathcal{E}} \mathbb{E} — anticipatory, introverted, consults stored experience. Driving (Feed-forward): d\mathcal{E}/d\tau = +\kappa \nabla_{\mathcal{E}} \mathbb{E} — reactive, extroverted, engages with immediate input. The same processor, the same meta-time \tau, the same spectral energy \mathbb{E}. Only the sign and the temporal reference differ.· A cognitive interpretation. Steering corresponds to introverted functions (Ni, Si, Ti, Fi): consulting internal memory, comparing against stored patterns, pausing before acting. Driving corresponds to extroverted functions (Ne, Se, Te, Fe): engaging with the external world, reacting to present stimuli, acting without hesitation.· Driving in neural networks. The reactive update does not compare against stored targets. It acts on the immediate input: W_{t+1} = W_t + \eta \nabla_W \mathcal{J}(W_t, x_t), where \mathcal{J} is an objective evaluated on the current input alone. This is necessary when there is no stored experience, when speed matters, when the environment is the teacher, and when generation is the goal.· Experiment 1: Reactive control (point mass). A Driving PD controller outperforms a Feed-backwards model-predictive controller on a simple positioning task (integrated error 0.72 vs 0.78, overshoot 8% vs 12%). When the environment is simple and predictable, consulting a model adds overhead without improving action.· Experiment 2: Sequence generation with lookahead. On character-level language modeling (Penn Treebank), Driving reduces perplexity from 78.2 (autoregressive) to 72.8 by reacting to incoherence as it emerges, rather than relying solely on patterns learned during training.· Experiment 3: Classifier-guided image generation. On CIFAR-10 diffusion models, Driving (adaptive guidance) achieves FID 4.05 and class accuracy 95.1%, outperforming fixed guidance (FID 4.31, accuracy 94.2%) by reacting to uncertainty as it arises during generation.· The combined architecture. The full processor operates in both modes simultaneously: d\mathcal{E}/d\tau = -\kappa_b \nabla_{\mathcal{E}} \mathbb{E}[\mathcal{E}_\tau] + \kappa_f \nabla_{\mathcal{E}} \mathbb{E}[\mathcal{E}_\tau^{\text{now}}]. The negative term anticipates (consults stored experience). The positive term reacts (engages with the present). The coupling constants \kappa_b and \kappa_f determine the balance. This is the architecture of cognition. Why this matters: The same processor, the same meta-time, only the direction changes. Steering anticipates. Driving acts. Both modes are necessary. Neither is sufficient alone. This is the mathematical framework for cognition itself: introversion consults the past; extroversion engages the present. The processor does both. Keywords: Driving, Feed-forward dynamics, Steering, Feed-backwards, equality processor, reactive control, sequence generation, diffusion guidance, introversion, extroversion, cognitive functions, Canvas Model, meta-time
Abstract Research on how non-natives process and learn binomials ( black and white ) is limited. The present study addresses this gap using online (eye-tracking) and offline (familiarity rating) tasks. Sixty non-native speakers of English (L1 = Arabic) read six stories seeded with 21 novel binomials in three conditions: one exposure, six exposures, and no exposure (i.e., only in post-test) in a counter-balanced design. Each item was also presented in the reversed order ( white and black ). The non-natives read the stories as their eye movements were monitored and answered comprehension questions. In addition to the novel binomials, 12 existing binomials (congruent with Arabic) were included in the passages as a baseline for comparison. After completing the reading task, the participants completed an offline rating task as a measure of declarative knowledge of the binomial configuration (i.e., word order). All items were rated twice, once in the forward direction and once in the reversed direction. Online results showed that non-natives were not sensitive to the configuration of existing binomials, and there was limited evidence of any sensitivity to novel binomials. Offline, non-natives showed sensitivity to the configuration restrictions of existing binomials but not novel ones.
Emotional memories persist within individuals and over generations. Past research has examined functions of parent-child memory sharing, but little work has assessed how the emotional qualities of memories are transmitted from parent to child and whether emotion transmission relates to memory content transmission. An understanding of how emotional memories are transmitted may be particularly important during adolescence, a developmental period marked by heightened sensitivity to emotional information, increasing independence from caregivers, and the emergence of mental health symptoms. The current study investigated the intergenerational transmission of parents’ autobiographical emotional memories to their teen offspring in healthy dyads using behavioral measures and natural language processing tools. We found that emotional valence and arousal linked to parents’ memories are transmitted from parent to teen. Greater parent-teen agreement in valence ratings was associated with greater parent-teen overlap in both subjective vividness ratings and objective memory content of individual memories. Subjective memory transmission was modestly related to lower mental health symptoms in teens after accounting for parent symptoms. These findings demonstrate that parents’ emotional memories are transmitted to their teens and provide preliminary evidence that autobiographical emotional memory transmission from parents could be a protective factor for mental health in adolescence.
The Canvas Model's equality processor operates in two complementary feed-modes. Feed-backwards (Steering) corrects errors via gradient descent: d\mathcal{E}/d\tau = -\kappa \nabla_{\mathcal{E}} \mathbb{E}. Feed-forward (Driving) anticipates goals via gradient ascent: d\mathcal{E}/d\tau = +\kappa \nabla_{\mathcal{E}} \mathbb{E}. The first four papers in this series explored Feed-backwards—training, convergence, regularization, and modularity. This paper explores Feed-forward. What this paper provides: · A formalization of anticipatory dynamics. The positive-sign dynamics enable the system to project forward, predict optimal trajectories, and act preemptively to maximize anticipated alignment with future goals. The update uses gradient ascent on a projected future state, not gradient descent on the current state.· Three experimental validations across domains: 1. Anticipatory continuous control (point mass positioning). Feed-forward Driving achieves integrated error of 0.72 \pm 0.11 vs 1.00 \pm 0.15 for reactive PD control and 0.78 \pm 0.12 for Model Predictive Control (MPC). Overshoot is reduced from 23% (PD) to 8% (Driving), beating MPC (12%). Driving matches MPC performance without a learned dynamics model or horizon optimization. 2. Lookahead for discrete sequence generation (character-level language modeling on Penn Treebank). Feed-forward Driving reduces test perplexity from 78.2 (autoregressive baseline) to 72.8 \pm 1.3. The lookahead projection anticipates grammatical constraints, avoiding locally probable but globally incoherent choices. Qualitative inspection confirms fewer repeated characters and more plausible word formations. 3. Adaptive guidance for image generation (classifier-guided diffusion on CIFAR-10). Standard classifier guidance uses a fixed scale w. Driving introduces adaptive per-step guidance: w_t = w_0 / (1 + p(y \mid x_t)). When the classifier is uncertain, guidance is stronger; when confident, guidance relaxes. Adaptive Driving achieves Fréchet Inception Distance (FID) of 4.05 \pm 0.14 vs 4.31 \pm 0.18 for standard guidance, and class accuracy of 95.1% vs 94.2% — better image quality and higher fidelity.· Connection to existing methods. Driving is not a new algorithm. It is a unifying principle that explains why Model Predictive Control, classifier-guided diffusion, lookahead optimizers, and RLHF work. All are manifestations of the Feed-forward mode, distinguished only by the choice of projection mechanism and the spectral energy being ascended.· The combined architecture. The full processor operates in both modes simultaneously: d\mathcal{E}/d\tau = -\kappa_b \nabla \mathbb{E}[\mathcal{E}_\tau] + \kappa_f \nabla \mathbb{E}[\mathcal{E}_\tau^{\text{proj}}]. The negative term corrects errors. The positive term anticipates goals. This is the mathematical framework for cognition itself: a system that both learns from its errors and acts on its anticipations.· Limitations acknowledged. Driving's effectiveness depends on the quality of the projection. If the projected future state is inaccurate, Driving can move toward the wrong target. The combined architecture (\kappa_b, \kappa_f > 0) mitigates this by allowing the system to both anticipate and correct. Why this matters: Steering corrects. Driving aims. The same processor, the same meta-time, the same spectral energy. Only the sign changes. The Canvas Model reveals that the fundamental dichotomy in AI—reactive vs. anticipatory, error-correcting vs. goal-seeking, introverted vs. extroverted—is not a philosophical distinction. It is a mathematical one: the sign of the update in meta-time. Keywords: Driving, Feed-forward dynamics, anticipatory control, lookahead generation, classifier guidance, diffusion models, gradient ascent, Model Predictive Control, RLHF, Canvas Model, equality processor, meta-time, Steering, Feed-backwards
Abstract This article examines a new direction in the digitalization of Kazakh linguistics–the development of a six-language parallel subcorpus based on literary texts. The research focuses on a multilingual database composed of texts in Kazakh, English, Turkish, Uzbek, Uyghur, and Azerbaijani. Its structural organization, metadata annotation system, and principles of paragraph-level alignment are described. The research material is based on a multilingual corpus compiled from Volume I of M. Auezov’s novel-epic The Path of Abai. The study provides a comparative analysis of syntactic, semantic, and pragmatic equivalence across translations in different languages. Particular attention is given to complex sentences, dialogic structures, forms of address and endearment, ethnomarked units, literary and poetic vocabulary, figurative expressions and idioms, as well as interjections. These elements are systematically classified, and their cross-linguistic equivalence levels are identified. The results demonstrate a high level of structural and semantic equivalence among Turkic languages, while English translations tend to prioritize functional and pragmatic adaptation. Furthermore, the six-language parallel subcorpus of the Kazakh language is shown to be a valuable resource for applied research in corpus linguistics, translation studies, comparative linguistics, and digital humanities. The scientific novelty of the study lies in the first-ever proposal of a six-language parallel subcorpus model in Kazakh linguistics based on literary texts, along with a comprehensive description of its metadata annotation, alignment, and analytical principles. The findings can be applied in corpus linguistics, translation studies, comparative linguistics, and the development of linguistic databases for artificial intelligence systems.
Affective word values have been widely studied across languages, often focusing on isolated words due to the difficulty of assessing emotionality in texts. This study examines whether written emotional content can be reliably captured using a specific software tool (Watson Natural Language Understanding). Thirty-three Spanish undergraduates wrote 150-word autobiographical texts in their L2 (English) before and after a training with emotional vocabulary. Normative valence ratings of content words obtained in the pre- and post-training phases were compared with sentiment scores generated by Watson NLU. Strong positive correlations were found between sentiment and normative valence scores in both phases, with stronger relations at post-training. Regression analyses confirmed that sentiment scores significantly predicted normative valence. Importantly, while normative valence did not differ between phases, sentiment scores increased after training. These results suggest that Watson NLU is a valid and sensitive tool for assessing emotionality in written language and its modulation through training.
The global deployment of large language models (LLMs) has raised concerns about cultural misalignment, yet the linguistic properties of fine-tuning datasets used for cultural adaptation remain poorly understood. We adopt a dataset-centric view of cultural alignment and ask which linguistic properties of fine-tuning data are associated with cultural performance, whether these properties are predictive prior to training, and how these effects vary across models. We compute lightweight linguistic, semantic, and structural metrics for Arabic, Chinese, and Japanese datasets and apply principal component analysis separately within each language. This design ensures that the resulting components capture variation among datasets written in the same language rather than differences between languages. The resulting components correspond to broadly interpretable axes related to semantic coherence, surface-level lexical and syntactic diversity, and lexical or structural richness, though their composition varies across languages. We fine-tune three major LLM families (LLaMA, Mistral, DeepSeek) and evaluate them on benchmarks of cultural knowledge, values, and norms. While PCA components correlate with downstream performance, these associations are strongly model-dependent. Through controlled subset interventions, we show that lexical-oriented components (PC3) are the most robust, yielding more consistent performance across models and benchmarks, whereas emphasizing semantic or diversity extremes (PC1-PC2) is often neutral or harmful.
We revisit punctuation-aware tree binarization for constituency parsing and ask whether dependency-induced headedness improves binary parser supervision. Although learned heads substantially outperform rule-based heads in intrinsic head prediction, they do not yield consistent parsing gains after debinarization. In particular, punctuation-conditioned evaluation shows that learned headedness underperforms rule-based binarization in macro-average punctuation-sensitive $F_1$, despite a small overall gain on CTB. Similar instability appears under cross-treebank transfer. These results suggest that \ycc{linguistically grounded} headedness is not necessarily parser-optimal when used as a binarization control signal. The paper presents a negative result: better head prediction does not imply better punctuation-sensitive constituency parsing.
In the context of globalization, digital communication, and the growing dominance of standardized language forms, dialectal lexical units are increasingly marginalized in everyday communication. However, dialects continue to function as vital carriers of national culture, historical memory, and social identity. This study aims to investigate the role of dialect-specific lexical units in reflecting national identity through a comparative analysis of Uzbek and English dialects. The research is grounded in linguocultural, sociolinguistic, and functional-semantic approaches. Empirical data were collected from regional Uzbek dialects and English dialectal sources, including spoken discourse, literary texts, and previous scholarly studies. More than fifty dialectal lexical units were selected based on their cultural specificity, emotional-evaluative potential, and limited equivalence in standard language. The data were analyzed using descriptive, comparative, and interpretative methods to reveal their semantic structure and communicative functions. The findings demonstrate that in Uzbek dialects, national identity is predominantly reflected through lexical units related to kinship (qudachilik), neighborhood relations (mahalla), and traditional labor practices. These units embody collective values such as social solidarity, mutual responsibility, and respect for tradition, and they possess strong emotional and evaluative connotations. In contrast, English dialects primarily reflect national identity through rural life vocabulary, occupational terminology, and class-based linguistic markers. Dialectal expressions in English often function as indicators of social class, occupational background, and individual identity, highlighting the role of social stratification in linguistic variation. The comparative analysis reveals that while both languages employ dialectal vocabulary as a means of cultural identification, Uzbek dialects emphasize collective and community-oriented identity, whereas English dialects foreground individual and class-based identity. The study concludes that dialect-specific lexical units should not be viewed as deviations from linguistic norms, but as essential components of national linguistic heritage. These findings have practical implications for linguocultural studies, language education, translation practice, and the preservation of intangible cultural heritage.
This paper introduces the first steps towards the creation of a novel resource for contemporary Sardinian within the Universal Dependencies framework. Sardinian is a Romance language spoken in Sardinia, an island belonging to the Italian Republic and located in the center of the western Mediterranean. It is a minority and endangered language, traditionally transmitted mainly orally, and characterized by a multiplicity of varieties (usually grouped into two macro-varieties Logudorese and Campidanese), all recognized as part of the Sardinian linguistic continuum. These varieties share basic morphosyntactic features, while presenting differences at the lexical level and in the realization of specific constructions. This internal variation can be particularly challenging with regard to the normalization of lemmas and the linguistic characterization of certain phenomena. The development of the treebank therefore aims to provide an annotated resource for contemporary Sardinian that takes into account the specificities of the different varieties, using Universal Dependencies to represent them within a unified theoretical framework, in order to facilitate both linguistic analysis and automatic processing. The present paper thus describes some linguistic characteristics of Sardinian and the attempts to encode them within the UD framework. Finally, we present the results of our evaluation of an NLP pipeline for Sardinian, trained on our corpus, for the Stanford Stanza parser.
This article analyzes the systemic crisis of Arabic culture from the 13th to 18th centuries and the pivotal role of Arab Christians in initiating the Arab Renaissance. The author explores the causes of stagnation in Muslim society, primarily the institutional dominance of taqlīd (imitation) and the widening gap between sacralized linguistic norms (fuṣḥa) and living speech. Central to the study is the scholarly contribution of the Maronite scholar Ibn Farhat, whose work bridged Western rationalism and Eastern tradition. Special focus is placed on his treatise «Baḥṯ аl-maṭālib wа ḥаṯṯ аl-ṭālib», which simplified Arabic pedagogy and integrated the language into the daily and liturgical practices of Christian communities. The paper emphasizes that the transition from Karshuni script to classical Arabic, alongside Ibn Farhat’s reforms, provided the ideological foundation for overcoming cultural isolation. It concludes that Lebanon’s Christian intellectuals, educated through European models like the Pontifical Maronite College, were the primary catalysts for modernizing Arabic philology and precursors to the 19th-century Enlightenment.
This paper examines how artificial intelligence (AI), machine learning algorithms, and automated digital systems shape linguistic practices, reinforce or challenge linguistic hierarchies, and influence communication in contemporary society. As digital platforms increasingly mediate human interaction, algorithms determine what content becomes visible, which linguistic varieties are privileged, and how users adapt their language to gain visibility and engagement. The study explores algorithmic bias in search engines, social media feeds, voice assistants, and automated moderation systems, highlighting how these technologies reproduce existing social inequalities related to class, caste, gender, and ethnicity. Drawing on sociolinguistic theories of language ideology, linguistic capital, and digital discourse, the paper argues that AI-driven communication environments are not neutral but deeply ideological. They shape linguistic norms, influence identity performance, and regulate public discourse. The findings underscore the need for critical sociolinguistic engagement with AI systems to ensure equitable, inclusive, and culturally sensitive digital communication.