Papers reviewed and determined not to be word norm studies. Use the flag icon to report errors or suggest re-inclusion.
18625 papers
The article is theoretical in nature. It focuses on the language awareness among young people. Language awareness enables code-switching and balancing between the sociolect known as the youthlanguage and the official, standard-compliant language. Depending on the context, language is usedby young people to identify themselves with the group and to manifest their independence, distinctiveness and uniqueness. It also reflects their integration into linguistic norms and their creative transformation of these norms. The level of language awareness determines which variety young people choose to use in particular situations. Drawing on the levels of language awareness identified in the literature, the author presents examples of both low and high levels of language awareness in thelinguistic competence of young people. The conclusion is presented within the context of the core curriculum, highlighting the need to correlate literary education with language education.
Recent technological advances have provided the chance to conduct Virtual Reality (VR) experiments with increased ecological validity, which in turn can elicit more naturalistic responses in immersed users. However, many studies still prefer highly controlled setups and passive stimulation, often due to the practical complexities in effectively associating cause (stimulus) to response in highly interactive and dynamic VR experiences. Many of these studies also rely on subjective ratings from participants recorded either after the experience (relying on memory) or during the experience (interrupting immersion). In this paper, we advance this experimental protocol in a large-scale feasibility study by (1) investigating affective changes in terms of valence and arousal ratings in various interactive 3D room-scale VR environments with (2) continuous valence and arousal self-ratings from a controller and (3) a novel wireless physiological facial EMG and PPG sensor setup specifically designed to record affect, without relying on memory or interrupting immersion. In this study, n=291 participants experienced neutral, positive and negative virtual environments in ’passive’ and ’active’ conditions. Continuous self-ratings and physiological measures confirmed the feasibility of detecting affective states in room-scale VR conditions. To our knowledge, this is the highest n in a feasibility study in affect detection to date. Our study generated the most populated physiological data library collected in VR, which also compares passive and active VR settings. This setup can provide a solid experimental foundation for VR affective computing studies in more unconstrained, ecologically valid environments.
Abstract Cognitive reappraisal is a fundamental emotion regulation strategy for mental and physical well-being, but how its neural mechanisms relate to individual differences remains poorly understood. In a consortium effort analyzing 40 fMRI datasets ( N =2,175), we examined the relationship between neural activation during reappraisal tasks and three core individual difference indices of reappraisal capabilities: (1) trait questionnaires, (2) task-based affective ratings, and (3) amygdala down-regulation. Strikingly, there was no shared overlap across these three common indices. Only a very weak correlation emerged between amygdala down-regulation and task-based affective ratings. Whole-brain analyses revealed no reliable neural associations with trait questionnaires, and associations with task-based affective ratings fell outside canonical emotion regulation networks (e.g., prefrontal circuitry). Moreover, amygdala down-regulation, often interpreted as a stable individual marker, was confounded by person-specific whole-brain responses — a limitation extending to fMRI research beyond the emotion regulation domain. These findings challenge the assumption that an individual’s prefrontal activity is a valid indicator of their reappraisal capabilities and suggest that common trait, behavioral, and neural measures might capture distinct facets of emotion regulation. More broadly, our results highlight concrete methodological challenges for fMRI research on individual differences, with implications extending beyond emotion regulation to the neuroscience of personality, psychopathology, and general well-being.
Idioms are undoubtedly important for second language (L2) learners, who encounter them in instructed learning, textbooks/resources and in out-of-class language use. While research on first language (L1) and L2 idiom comprehension shows how well L1/L2 speakers understand various idioms and the role of different predictors, important questions remain about how knowledge varies with more difficult task types and stimuli, how well L1 ‘norms’ serve L2 learners, how subjective and objective predictors of idiom knowledge interact and how L2 learner inferencing works in learning idioms. To address these issues, university-age L1 and L2 English (L1 German) participants provided meaning descriptions and familiarity ratings for 100 challenging idioms from learner resources, and each idiom was assigned an OpenAI-generated transparency rating, corpus-based frequency and to one of six cross-language overlap (CLO) types. Descriptive statistics showed lower and more varied idiom meaning knowledge than might be expected, especially for the L1ers, who were some way off ceiling level. Mixed-effects regression revealed familiarity and transparency as positive L1 and L2 knowledge predictors, but groups differed in sensitivity to idiom frequency, which only mattered for the L1ers and CLO, which (as expected) only mattered for the L2ers, who mistook false friends as genuine allies.
Characteristics of emotion, such as valence and arousal can be evaluated using self-reported affective ratings and electroencephalogram (EEG) to gain a better understanding of individual differences in diversified populations. The International Affective Picture System (IAPS) and AI-generated counterparts were used to elicit emotional responses that were collected from the self-assessment manakin (SAM) rating scale for valence and arousal. EEG data were used to observe biomarkers of emotional processes related to the presentation of these stimuli. These methods were correlated to individual differences such as sex and depression and anxiety related questionaries. The study showed significant sex differences in the self-reported affective ratings, where females showed greater aversive affect compared to males. EEG data showed that there was less alpha reduction relative to baseline in percent for individuals who scored high on the BDI-II, which suggests biomarkers of emotional dysregulation, such as anhedonia. Overall, the study highlights the differences in emotional processes for variable populations, which has implications for intervention and targeted treatment efforts.
This study examines the intersection of language, identity, and politics in Nigerian online discussion through a sociolinguistic analysis of Nairaland, Nigeria's largest internet forum. Drawing on Communication Accommodation Theory (CAT) and sociolinguistic theory, the research investigates how participants strategically use code-switching, Nigerian Pidgin, and local languages to negotiate identity, voice dissent, and establish community in politicised debates. Three selected threads, "Igbo Founded Ile-Ife," "Okrika Women Protest Unlawful Suspension of Fubara," and "Rivers Emergency Rule", serve as case studies for the analysis of how language works as a vehicle to convey ethnic pride, political grievance, and gendered opinion in online environments.Participants are observed to employ convergence, divergence, and maintenance strategies to show group membership, claim differences, or refuse linguistic norms. Nairaland's language options are not arbitrary but serve affective, rhetorical, and ideological functions, permitting speakers to negotiate group membership or sociopolitical grievances. The study reveals how Pidgin English and blended varieties of language are employed to construct a collective sense of identity and solidarity, particularly in protest or satirical contexts. It also discloses the gendered politics of exclusion whereby women's political voices are diminished through sexist and class-based labelling, an outcome of more profound societal biases replicated within the virtual spaces.The study contributes to the knowledge of digital multilingualism and bottom-up creativity in African online forums, highlighting the importance of ethical digital literacy and the moderation of inclusive discourses. By examining language use on Nairaland, the book enlightens readers on how Nigerians imaginatively appropriate linguistic material to shape power relations, assert their identities, and construct solidarity in the digital sphere. Relevant to scholars in sociolinguistics, digital humanities, and African studies, the book presents a case study for examining political discourse and identity creation in multilingually complex cyberspaces.
This paper presents an approach to integrating Latin inflected forms and corpus attestations within a Linked Open Data (LOD) framework, enhancing interoperability between Wikidata and the LiLa knowledge base. Building on the PrinParLat lexicon of Latin verb principal parts, we generate the complete set of inflected forms for over 8,000 verbs, encoded as RDF in a dedicated Wikibase instance. These forms are linked to the Index Thomisticus Treebank (ITTB), whose morphologically annotated tokens are related to corresponding forms based on segmental identity, lemma alignment, and mapped morphological features. Our generation and linking process achieves over 95% coverage of ITTB verbal tokens, demonstrating the robustness of our pipeline even for Medieval Latin data. By aligning Paralex, Wikidata, and LiLa ontologies, we ensure semantic interoperability and facilitate future integration into Wikidata. Beyond Latin, this workflow provides a reproducible model for linking inflectional paradigms and corpus attestations in other languages.
Pshal P’shaw investigates the sonic instability of speech—where phonetic dissonance, vocal fragmentation, and gestural sound challenge structured linguistic norms. Developed during my residency at the Max Planck Institute for Empirical Aesthetics, this work explores how speech patterns morph in response to unfamiliar dialects, sonic environments, and computational intervention. Inspired by Hermann Finsterlin’s fluid, dimensionally unbound architectural forms, I approach language as an unfixed structure, revealing phonetic articulation's unstable, adaptive nature. At its core, Pshal P’shaw is a multi-channel sculptural sound installation, neural instrument, and experimental phonetic composition that deconstructs oral adaptation and linguistic convergence. The project is based on 28 recorded voices engaging with a script designed to test the phonetic contours of a contemporary US Western dialect. Participants articulated words in controlled EEG lab settings, allowing for an analysis of phonetic deviation, mimicry, and emergent vocal patterns. I collaborated with Cornelius Abel, the Head of the EEG laboratory at the Max Planck Institute for Empirical Aesthetics, to create a participant-based study. This study focuses on neural responses to phonetic adaptation and linguistic instability. It examines how the brain processes disrupted phonetics, prosodic shifts, and unfamiliar articulatory patterns, providing insights into cognitive processing beyond structured linguistic norms. By positioning phonetic instability as an expressive and cognitive site, this work aligns with my broader research on interstitial sonic spaces—the liminal areas between signal and noise, order and disruption, articulation and abstraction. The project extends into machine learning and human-machine interaction, where a neural algorithm dissects and reorganizes speech based on phonetic similarity and difference. The system works within a granular information base of over 3,000 phonetic variations, revealing the rhythmic, percussive, and spectral structures embedded in vocal articulation. Working with designer and Max/MSP specialist Matthew Ostrowski, I developed a custom Max/MSP script to explore how an AI-driven phonetic instrument can exist in a wandering state rather than optimization. Through sessions envisioning a system that resists linear progression, we designed a model that moves fluidly between pattern recognition and deviation, continually reorienting itself within the dataset rather than seeking a fixed resolution. This keeps the AI suspended in a process of recombination, allowing for unexpected phonetic proximities, spectral collisions, and recursive vocal transformations. The work employs Flucoma and MUBI as Max/MSP patches for phonetic and vocal analysis, mapping articulatory gestures through spectral decomposition, transient detection, and dynamic filtering. The script was developed in MATLAB by the EEG labs, enabling phonetic feature extraction and computational manipulation. A 12-channel sculptural sound installation hosts the generative algorithm, continuously reshaping the recordings based on vocal proximity and divergence. Simultaneously, a 5-channel raster video system maps the frequency data of each voice onto facial deconstructions, amplifying the inherent instability of articulation. A special edition record release with an accompanying booklet is available through Raster Media (Germany). This vinyl expands the project’s inquiries into a fixed yet fragmented format, extending the phonetic research into a sonic document. Side A, “Transients Script,” isolates the percussive elements of speech, revealing the microtonal instability of articulation as a rhythmic force. Side B examines spectral interplay across the 28 voices, exposing harmonic shifts, tonal variances, and dynamic resonances. Pshal P’shaw interrogates how language operates beyond fixed linguistic hierarchies, where vocal adaptation and sonic fragmentation become sites of expanded meaning-making by engaging with phonetic mimicry, algorithmic processing, and embodied sound. In doing so, the work contributes to larger conversations in experimental phonetics, accessibility in assistive sound technologies, and the role of noise in cognitive and communicative systems. Through historical research (on linguistic shifts in 19th-century US boomtowns) and contemporary computational modeling, this project reimagines how phonetic structures evolve, break apart, and reform—not as errors but as integral components of human interaction and cognition. Notes on Contributor Victoria Keddie is an artist whose work explores sound beyond structured language, investigating how communication fractures and reconstitutes through error, interference, and technological mediation. She has been Co-Director of E.S.P. TV for over a decade, using televisual media for performance. Her work has been performed and exhibited internationally, with recent fellowships at the NYSCA/NYFA, the Max Planck Institute for Empirical Aesthetics, Elektronmusikstudion, and the Bemis Center for Contemporary Art. Keddie’s video works are distributed by Lightcone (FR) and The Filmmakers Co-op (US), and her sound works are released with Raster Media (DE), Chaikin Records (US), and Fridman Gallery (NYC/US).
Nyakyusa is the Bantu language registered as M.31 and it is spoken in the South-Western part of Tanzania, particularly in Mbeya Region and some parts of Njombe. Similar to other languages, the Nyakyusa is in contact with the English language whose lexicon is fed by the English loanwords. This study aimed to explore how English loanwords are adapted and integrated into the Nyakyusa language. English is one of the most widely spoken languages globally, and its contact with Nyakyusa has become the primary source of loanwords, with approximately 95 percent of Nyakyusa’s borrowed vocabulary originating from English. The English loanwords in Nyakyusa are adapted to suit the Nyakyusa morphophonotactics because the two languages have dissimilar morphophonotactics. Additionally, English has influenced the phonological structure of Nyakyusa. This is because borrowed words are adapted to fit the sound system of the Nyakyusa language. The theoretical framework for this study is based on the Lexical Phonology and Morphology (LPM) theory. The study adopted a descriptive research design, structured within an interpretive research paradigm, and employed a qualitative research methodology. The study employed a purposive sampling of 14 native Nyakyusa speakers. Data were collected through participant observations, interviews, and documentary reviews and subsequently analysed thematically. The findings revealed that the Nyakyusa language integrates English loanwords through several morphophonological processes, including segmental substitution, vowel epenthesis, glide epenthesis, segmental deletion, and segmental retention. These processes help to modify English morphophonotactic elements that violate the Nyakyusa language norms. The researchers recommend conducting comparative studies on the nativisation processes in other Bantu languages in order to contextualise the findings and highlight unique or shared features within the language family.
This chapter focuses on the ways in which maskilic (Jewish enlightenment) authors reshaped elements of the Hebrew language in order to make it suitable for use as a vehicle of modern European-style literature. Maskilic authors were motivated by the principle that the creation of a Hebrew literary canon based on European-language models was an essential component of the Jews’ Enlightenment. They developed a wide range of new written genres which had not previously existed in Hebrew, including original and translated novels, short stories, novellas, anecdotes, feuilletons, and plays; a flourishing press; popular science (e.g., astronomy, physics, biology, chemistry, zoology), politics, geography, and history; and fiction and nonfiction works for young readers. This large-scale endeavor resulted in the emergence of numerous linguistic innovations. In this chapter we examine several key yet understudied innovative features of maskilic Hebrew: 1) realistic colloquial dialogue, including a) new punctuation norms, b) use of interjections, and c) forms of address; 2) discussions of contemporary scientific and political topics using newly developed Hebrew vocabulary and European loanwords; and 3) prose syntax based on German and Russian stylistic models. These new components of Hebrew paved the way for the later vernacularization of the language in Ottoman Palestine.
This article investigates the pragmatic functions of first-person pronouns (I/we) in political speeches and formal writing across multiple language systems. By analyzing the discourse of prominent political leaders such as Shavkat Mirziyoyev, Joe Biden, and Emmanuel Macron, the study explores how first-person pronouns function to construct authority, inclusiveness, and responsibility. The research highlights variation across languages in terms of politeness, formality, and rhetorical strategy. Methodologically, the study employs comparative discourse analysis and pragmatic interpretation of political and academic texts. The findings demonstrate that the usage of "I" versus "we" reflects not only linguistic norms but also culturally embedded leadership styles.
The escalation of socio-economic and geopolitical tensions is heightening the potential for conflict not only in intercultural but also in monocultural communication. Within this context, the media plays a significant role in shaping public sentiment and fostering situations conducive to conflict. This study aims to define the Conflictogenicity Index of Media Discourse (CIMD), examine its correlation with national communicative styles, and analyse the specific realisation of Lexical Pragmatic Markers of Conflictogenicity (LPMCs) in contemporary Mexican and American media. A particular focus is placed on devising a methodology for calculating the CIMD and trialing it through the analysis of empirical data. The analysis draws on a corpus of 680 online publications from 2021 to 2025, sourced from outlets including La Jornada, Excélsior, El Milenio, Fox News, The New York Times, The New York Post, etc. These publications address the migration crisis − an issue of common concern to both Mexico and the United States − thereby framing the investigation of conflictogenicity within an “Us vs. Them” dynamic. The findings confirm that norms of national communicative style govern the selection, frequency, density, and permissible strength of negative evaluation in LPMCs. This study also reveals that both the perceived intensity of a conflictogenic marker and the overall CIMD are subject to cross-cultural variation. To illustrate, articles deemed highly conflictogenic within Mexican linguoculture may be rated as medium or even low on the CIMD scale compared to American media. Ultimately, the research demonstrates that LPMCs, as key indicators of conflictogenicity in discourse, not only signal underlying social tensions but also help to pinpoint the society’s most pressing issues.
Effective and safe human-machine collaboration requires the regulated and meaningful exchange of emotions between humans and artificial intelligence (AI). Current AI systems based on large language models (LLMs) can provide feedback that makes people feel heard. Yet it remains unclear whether LLMs represent emotion in language as humans do, or whether and how the emotional tone of their output can be controlled. We assess human-AI emotional alignment across linguistic-cultural groups and model-families, using interpretable LLM features translated from concept-sets for over twenty nuanced emotion categories (including six basic emotions). Our analyses reveal that LLM-derived emotion spaces are structurally congruent with human perception, underpinned by the fundamental affective dimensions of valence and arousal. Furthermore, these emotion-related features also accurately predict large-scale behavioural data on word ratings along these two core dimensions, reflecting both universal and language-specific patterns. Finally, by leveraging steering vectors derived solely from human-centric emotion concepts, we show that model expressions can be stably and naturally modulated across distinct emotion categories, which provides causal evidence that human emotion concepts can be used to systematically induce LLMs to produce corresponding affective states when conveying content. These findings suggest AI not only shares emotional representations with humans but its affective outputs can be precisely guided using psychologically grounded emotion concepts.
The present dissertation examines the role of context-specific simulations in influencing the complexity of affective experiences, drawing on a constructivist approach to emotion. To link literatures on mental simulation and emotion, in Chapter 1 a connection is made through the grounded theories of cognition. Chapter 2 describes the development of a novel dataset consisting of context-dependent stimuli (i.e., 1,381 picture-word cues derived from 320 pictures-only stimuli) validated through online experiments (NExp1= 1,934; NExp2 = 403). Hence, an investigation of how contextual information influences the affective experience is illustrated, revealing that context more often enhances affective complexity by widening, rather than narrowing, the variation in the between-subject valence ratings. Chapter 3 employs a set of stimuli selected from Chapter 2 in a lab-based experiment in which participants (N = 30) rated affect intensity, and reported which emotions and bodily sensations experienced in response to generating both mental images and verbal thoughts. Mental imagery was found to enhance emotion complexity as reflected in the richness of the reports provided, with affect intensity and autobiographical recall accounting for the effect. Finally, in Chapter 4 the influence of mental imagery on emotion complexity will be studied across the imagery spectrum. To this end, in an online experiment participants (N = 72) completed measures of imagery vividness, alexithymia and gave written reports on how they feel when experiencing emotions at varying levels of valence and arousal, to obtain indexes on the complexity of emotion conceptualization. In line with the predominant literature, the more vivid the visual mental imagery of participants, the less alexithymia was reported, i.e., the less impaired is the process of emotion conceptualization. As highlighted in the final chapter (Chapter 5), overall the present dissertation contributes to deepening the study of the relationship between mental imagery and emotion. Assuming variation as inherent to emotion, consistently through different experimental designs, methods, languages and indexes, it is shown how context-specific simulations enrich the emotional sphere, by enhancing the complexity of affective experiences.
This paper examines how an Indonesian live-streamer and his audiences create a hybrid communal identity by using vulgar language and code-mixing with reference to the popular YouTube livestream @deandeankt. With the digital medium transforming the world into a global village, the streaming culture in Indonesia provides a distinct perspective into how localized linguistic practices interact with global trends in the internet. The interpretivist-constructivist paradigm applied in the current case study research was aimed at understanding the way language and interaction produced hybrid identities. It gathered linguistic and paralinguistic data of @deadndeakt’s livestream for 90 minutes with netnography and thematic analyses. According to the study, abusive swearing (e.g., "kontol [dick]", "goblok [stupid]") represents 89.3 percent of the vulgar language and serves as playful insults, which enhance group cohesion. Emphatic swearing ("Anjing! Gua kalah! [Bitch! I lost!]") and cathartic swearing (“Ngentot, mic-nya rusak! [Fuck, the microphone is broken!]”) increase the emotional involvement. Intra-sentential blends, especially code-mixing (“Brightness-nya kita bikin tiga kali [We increase the brightness threefold]”), prevails in 85 percent. Other examples combine English gaming slang (template, unarchive) with Indonesian structure and local dialects, like a Javanese swear word (“Cok [Damn]”). Pronunciations, such as BTW [read in Indonesian: be te we] are even more localized global terms. It is analyzed that the vulgar language usage assists in persona creation and building of relationships with the audience, creating a tight-knit, exclusive community in which such language represents belonging and emotional release, despite its offensiveness in the larger society. Moreover, streamers combine Indonesian and English with gaming jargon to successfully appeal to their bilingual audience, so that they could be technically accurate and still fit into the global gaming cultures. These activities form a hybrid identity, where speakers possess a strong sense of Indonesian culture but are consistent with international trends, which strengthens group exclusivity. The results draw attention to live-streaming as a contact zone in which language negotiates belonging, which disrupts strict cultural boundaries. The study highlights the sociolinguistic innovativeness of online communities and recommends future research on offline effects of these online linguistic norms and the effects of the hybrid identities and communicative practices used in live-streaming communities on daily language, social interaction, and identity formation in the real world. The authors also suggest cross-cultural comparative research to find out whether the phenomena are specific to the Indonesian situation or have they become a global trend.
Exegesis is a pivotal topic in linguistic studies, explored across multiple levels: phonetic, morphological, syntactic, and semantic, in addition to its connections with other disciplines.This study focuses on exegesis at the morphological level, aiming to define its nature by examining its causes from the perspectives of both the speaker and the listener.It also seeks to establish essential guidelines, document its manifestations in nouns and verbs, and detail the methodologies employed in its analysis.The study's significance lies in its contribution to enhancing the understanding and construction of speech, as well as comprehending the linguistic norms adhered to by Arabs in their expressions.
This project contains two lexical databases of Italian blends collected using different methods. Both resources have been analysed in my Ph.D. thesis "Un itangliano semplesso. Analisi di due campioni di parole macedonia", and are also the core data of other publications.
Background Lexical retrieval impairments are a hallmark feature of Primary Progressive Aphasia (PPA) and play a central role in diagnosis and clinical management. Lexical retrieval is a multiple-component process, and impairment to each of the components of this process gives rise to a different pattern of deficit. However, most current research on lexical retrieval in PPA categorizes deficits only broadly as either semantic or phonological, without identifying the exact functional locus of impairment or accounting for finer dissociations within these broad classifications.Aims The current study applies a cognitive neuropsychological model of lexical retrieval and a corresponding classification framework to systematically categorize lexical retrieval impairments in 40 Hebrew-speakers with PPA.Methods We used a comprehensive assessment battery that includes picture-naming, word-picture matching, picture association, and nonword repetition, and conducted a detailed error analysis. The classification algorithm considers success rate in tasks compared to age-matched norms and error types for each participant.Results We identified lexical retrieval impairments in 75% of the participants. Selective impairments were found in each of the five cognitive components involved in lexical retrieval, with some participants diagnosed with more than one impaired component. The impaired components include the conceptual system (5 participants), the semantic lexicon (13 participants), the phonological output lexicon or the connection between semantic and phonological lexicons (18 participants), and the phonological output buffer (19 participants). Each impairment was associated with a distinct pattern of errors in both naming and in the additional language assessments. The findings reveal dissociations between different types of lexical retrieval impairments, reflecting the modular nature of lexical retrieval.Conclusions Lexical retrieval deficits in PPA are prevalent and diverse. Capturing this diversity requires a componential, theory-driven model and classification framework. This framework more accurately describes the functional locus of impairment compared to the variant-based classification and allows for patients to be diagnosed with more than one type of impairment. Accurate diagnosis and classification are essential for effective treatment and the development of communication strategies for patients with PPA.
Este é um estudo em andamento sobre Processamento de Língua Natural de um corpus de Narrativas Clínicas em português brasileiro com duas versões anotadas: uma pela máquina e outra por humanos. A frequência dos rótulos das classes de palavras e das relações de dependência dos tokens em cada versão é calculada e uma análise guiada pelo corpus é realizada, destacando as correções feitas pelos humanos nas anotações da máquina. A comparação dessas anotações permite a criação de treebanks que podem ser usados para treinar novos modelos usando técnicas de aprendizado de máquina e para aprimorar diversas aplicações de Processamento de Língua Natural com corpus da área biomédica. Além disso, essa comparação permite a análise da consistência teórica de anotação, a fim de identificar o sistema gramatical desse tipo de corpus e criar guias de anotação para Narrativas Clínicas em português brasileiro de acordo com as Dependências Universais.
The article presents the results of a cross-cultural affective images perception study by Americans and Russians and reveals the degree of cultural factor influence on the stimuli assessment by American and Russian men and women. The hypothesis is that assessments of affective images by American and Russian respondents will have statistical differences due to the linguistic and cultural specificity of the ethnic groups; it is also assumed there are cross-cultural gender differences in the assessment. The study used the method of psycholinguistic questioning with seven-point scaling. 84 images from the open American database of affective images (“Open Affective Standardized Image Set”) were used as research material. The respondents were 34 men and 58 women. The results of the analysis did not show significant cross-cultural differences in ratings of affective images with reference to valence type or emotional evaluation/response. In general, Americans and Russians had a similar distribution of image ratings. However, a statistically significant difference has been found in the ratings of images with different valence types (P < 0.001). Negative and positive images were rated higher by Russians in terms of emotional evaluation, in contrast to Americans, most of whose emotional responses had neutral ratings. There was also a statistically significant difference in the ratings of different thematic images (P < 0.05). Nature images were rated by Russians as causing a feeling of comfort, while Americans noted their neutral impact on them. Images of objects, on the contrary, received the opposite ratings from the respondents. Moreover, cross-cultural gender differences have been revealed between Russian and American women in image ratings based on emotional evaluation and valence parameters (P < 0.05). Russian women rated most of the images as having a positive or negative impact, while the majority of American women’s ratings tended to be neutral. This confirms the influence of the emotional stimulus, valence type, image theme, as well as gender factor on the processing of emotionally coloured units by representatives of different cultures.
The study examines how children, parents and staff in a kindergarten create a social space in the kindergarten’s cloakroom through (linguistic) actions and language choices. Based on one year of ethnographic fieldwork, which includes participant observation, documentation of the kindergarten’s semiotic landscape, field conversations and interviews with staff, the study shows how the cloakroom becomes a site for multilingual practices, while the kindergarten otherwise is dominated by a Norwegian language norm. The study demonstrates how children and adults take ownership of this space and create a multilingual environment through their actions. The analysis is grounded in Lefebvre’s theory of the production of social space – through spatial practice, representations of space and lived space – Gadamer’s perspectives on play and Pascual-de-Sans’ concept of idiotopy. The findings reveal that language choices and the construction of the cloakroom as a social space are influenced by complex processes related to place, time and agency. When the cloakroom is less in focus for the staff, it can become an important site for children’s play, where they negotiate both linguistic norms and other rules. The study argues for the significance of such social spaces as part of the kindergarten’s linguistic environment, where children can take ownership and make language choices that deviate from established majority language norms. At the same time, the study highlights the importance of time in research on language and place, both theoretically and methodologically, as well as the roles of researchers in gaining access to such spaces through invitations from the children.
This chapter explores the challenges of English language teaching in Brazil, where the language is often linked to colonialism, capitalism, and patriarchy. It introduces the Afro-Brazilian concept of the pedagogy of encruzilhadas, inspired by Exu, which embraces multiplicity, transformation, and resistance to epistemic colonialism. By challenging traditional linguistic norms and advocating for languaging as a dynamic, situated practice, this chapter argues for a decolonial approach to language teacher education. It highlights how encruzilhadas foster pedagogical dissensus, enabling educators to navigate tensions between colonial legacies and local knowledge systems, ultimately reimagining language education beyond imposed Eurocentric frameworks.
EMOSIC: Emotional Motion and Music Dataset Overview The **EMOSIC Dataset** features over **7.5 hours** of high-precision, synchronized audio and upper-body motion capture data. It captures the performances of **10 professional musicians** from a top-tier conservatory playing **traditional Chinese plucked instruments** (Zhong Ruan, Liuqin, Soprano Ruan, and Bass Ruan). This dataset provides a valuable **non-Eurocentric resource** to the affective computing and human-robot interaction communities. It is specifically designed to facilitate research into: - The **"Dominant-Amplifier"** mechanism in multimodal emotion perception. - Cross-cultural generalizability of emotion recognition models. - Generative motion synthesis for naturalistic robotic behaviors. Key Features - 90 Hz optical motion capture (15-marker upper-body skeletal model). - 16-bit, 16.0 kHz synchronized recordings. - Focuses on non-Western musical instrumentation and expressive gestures. --- Dataset Organization The dataset is organized into 3 main directories: ### 1. Data The complete dataset containing all **3,388 unique 8-second clips**. * **bvh/**: Processed 3D motion capture data in Biovision Hierarchy (`.bvh`) format. * **wav/**: Corresponding audio recordings in Waveform Audio File (`.wav`) format (16-bit, 16.0 kHz, mono). ### 2. Stimuli The subset of **60 clips** used as stimuli in the perceptual validation experiment described in the associated paper. These are partitioned into three modality conditions and three counterbalanced sets. * **A/**: Audio-Only clips (`.wav`). * **V/**: Visual-Only clips, rendered as silent skeletal animation videos (`.mp4`). * **AV/**: Audio-Visual clips, with synchronized audio and skeletal animations (`.mp4`). ### 3. Validation Contains raw data and metadata from the perceptual user study ($N=30$), provided in tabular (`.csv`) format. * **annotations_A.csv**: Raw Valence-Arousal ratings for Audio-Only condition. * **annotations_V.csv**: Raw Valence-Arousal ratings for Visual-Only condition. * **annotations_AV.csv**: Raw Valence-Arousal ratings for Audio-Visual condition. * **demographics.csv**: Participant demographic info (age, gender, musical background), linked by unique ID. * **participants.csv**: Counterbalanced group assignment for each participant. * **post_survey.csv**: Post-experiment qualitative feedback. --- Dataset Structure Once unzipped, the directory structure is as follows: ```text EmosicDataset_v1.0/ ├── README.md ├── Data/ │ ├── bvh/ # 3,388.bvh files │ │ ├── 0000.bvh │ │ └──... │ └── wav/ # 3,388.wav files │ ├── 0000.wav │ └──... ├── Stimuli/ # 60 clips used in user study │ ├── A/ │ │ ├── set1/ │ │ ├── set2/ │ │ └── set3/ │ ├── AV/... │ └── V/... └── Validation/ # User study results ├── annotations_A.csv ├──... └── post_survey.csv ``` File Naming Convention The core data files in `Data/bvh` and `Data/wav` use a consistent 4-digit numerical identifier (from `0000` to `3387`). * **Example:** `0000.bvh` is the motion data corresponding strictly to `0000.wav`. Files within the `Stimuli` folders utilize the same identifiers, allowing easy mapping back to the full dataset. Citation If you use this dataset in your research, please cite the following paper: > **Asymmetrical Emotional Synergy between Music and Body Motion: Insights from the EMOSIC Musical Performance Dataset** > *Authors: Kaiyuan Ma, Biyun Zhang, Jun He* > *Scientific Reports, Under review* Contact For questions regarding the dataset, please contact: **Jun He** (jhe@sjtu.edu.cn) School of Mechanical Engineering, Shanghai Jiao Tong University
This paper presents an extended statistical parsing framework for Tree-Adjoining Grammar (TAG) that incorporates part-of-speech (POS) information to enhance syntactic disambiguation, improve accuracy, and increase cross-lingual adaptability. While TAG provides a linguistically expressive mechanism for representing complex syntactic phenomena such as recursion and long-distance dependencies, however, conventional statistical TAG parsers remain largely constrained by their reliance on lexical anchors, which limits generalization across languages and leads to inefficiencies in ambiguous contexts. To address this, we improvise the statistical TAG formalism by conditioning derivation decisions on both lexical items and their associated POS tags, thereby enriching the feature space with syntactic category information. Beyond the baseline framework, this extended version introduces three major contributions. First, it integrates POS-based features into both generative and discriminative models, enabling robust handling of unseen or low-frequency lexical items. Second, it presents a cross-lingual evaluation using multilingual treebanks covering English to Indian language pairs, demonstrating consistent improvements in parsing accuracy and a 40–45% reduction in parsing time compared to conventional lexicalized TAG parser. Third, it provides an expanded analysis of computational efficiency, error patterns, and scalability across varying sentence lengths and linguistic families. Experimental results on a dataset of 15,000 annotated sentences reveal that the latest parser achieves significant gains in both accuracy and efficiency, with stable performance even in low-resource scenarios. The framework’s design further allows integration with neural embeddings, opening pathways toward hybrid symbolic–neural parsing models. Overall, the proposed POS enriched cross-lingual TAG framework offers a scalable, linguistically grounded, and computationally efficient solution for modern Natural Language Processing (NLP) tasks, including machine translation, information extraction, and question answering.
Large language models (LLM) perform outstandingly in various downstream tasks.However, there is limited understanding regarding how these models internalize linguistic knowledge, so various linguistic benchmarks have recently been proposed to facilitate syntactic evaluation of language models (LM) across languages.This paper introduces QFrCoLA (Quebec-French Corpus of Linguistic Acceptability Judgments), a normative binary acceptability judgments dataset comprising 25,153 in-domain and 2,675 out-of-domain sentences.Our study leverages the QFrCoLA dataset and seven other linguistic binary acceptability judgments corpus to benchmark eight LM.The results demonstrate that, on average, finetuned Transformer-based LM are strong baselines for most languages and that zero-shot binary classification LLM perform worse than the naive baseline on the task.However, for the QFrCoLA benchmark, on average, a finetuned Transformer-based LM outperformed other methods tested.It also shows that pretrained cross-lingual LLMs selected for our experimentation do not seem to have acquired linguistic judgment capabilities during their pre-training for Quebec French.Finally, our experiment results on QFrCoLA show that our dataset, built from examples that illustrate linguistic norms rather than speakers' feelings, is similar to linguistic acceptability judgment; it is a challenging dataset that can benchmark LM on their linguistic judgment capabilities.
Sensory perception and emotional evaluation of a stimulus are deeply interconnected with sensory awareness and attention. Mindfulness meditation is a well-established tool for enhancing attention and awareness, but can this heightened awareness influence how we perceive and emotionally evaluate tactile stimuli? In this work, we investigate how mindfulness meditation influences sensory perception and emotional evaluation of gentle stroking stimuli delivered to the forearm using ultrasound-based mid-air haptics, providing tactile sensations without direct skin contact. Two studies were conducted to assess the effects of mindfulness on (1) perceived intensity, confidence, valence, and arousal in response to simple linear strokes and (2) pattern recognition and emotional evaluation of complex mid-air haptic patterns. Our findings suggest that mindfulness enhances perceived intensity, confidence, and valence ratings, but does not influence arousal ratings. Pattern recognition was challenging for the participants on the forearm leading to poor accuracy but mindfulness still enhances the recognition. These findings highlight the potential of mindfulness-enhanced haptic interactions in digital wellbeing and virtual environments. Future research should explore long-term mindfulness training, multi-sensory integration, and adaptive haptic systems to further optimize affective haptic experiences.
Music serves as a powerful tool for emotion regulation, particularly in adolescents, who experience emotional challenges. Understanding the determinants shaping their perception of musical emotions may help optimize music-based interventions, especially for those with ADHD. This online study examined how familiarity, musical affinity, and ADHD diagnosis influence adolescents' judgments of musical excerpts in terms of arousal and emotional valence. A total of 138 adolescents (38 ADHD, 100 controls) rated 55 excerpts for arousal, valence, and familiarity using 10-point Likert scales. Musical affinity was conceptualized as a multidimensional construct encompassing musical experience, listening diversity, and receptivity to musical emotions. A cluster analysis identified two affinity profiles (low and high), and ANCOVAs tested the effects of affinity, ADHD, and familiarity on arousal and valence judgments. Familiarity strongly affected both arousal and valence. High-affinity adolescents judged excerpts as more pleasant and familiar, though arousal ratings did not differ between affinity profiles. Familiarity effects on emotional valence were stronger among lower-affinity adolescents. ADHD status did not significantly affect ratings. Overall, the study underscores music's potential for emotion regulation and its relevance in educational, clinical, and self-care contexts.
Code-switching presents a complex challenge for syntactic analysis, especially in low-resource language settings where annotated data is scarce. While recent work has explored the use of large language models (LLMs) for sequence-level tagging, few approaches systematically investigate how well these models capture syntactic structure in code-switched contexts. Moreover, existing parsers trained on monolingual treebanks often fail to generalize to multilingual and mixed-language input. To address this gap, we introduce the BiLingua Parser, an LLM-based annotation pipeline designed to produce Universal Dependencies (UD) annotations for code-switched text. First, we develop a prompt-based framework for Spanish-English and Spanish-Guaraní data, combining few-shot LLM prompting with expert review. Second, we release two annotated datasets, including the first Spanish-Guaraní UD-parsed corpus. Third, we conduct a detailed syntactic analysis of switch points across language pairs and communicative contexts. Experimental results show that BiLingua Parser achieves up to 95.29% LAS after expert revision, significantly outperforming prior baselines and multilingual parsers. These results show that LLMs, when carefully guided, can serve as practical tools for bootstrapping syntactic resources in under-resourced, code-switched environments. Data and source code are available at https://github.com/N3mika/ParsingProject
Prominent U.S. legal cases have turned on the resolution of linguistic ambiguities that arise from interactions between negation and coordination. Negative disjunction sentences (ND, “John didn’t buy cake or cookies”) are ambiguous between two readings: a neither-nor reading and a not-both reading. Negative conjunction (NC, “John didn’t buy cake and cookies”) is similarly ambiguous. Experimental findings on such scope ambiguities (Jasbi et al. 2023, Tobia et al. 2023) suggest that in general, listeners favor the neither-nor reading for ND but are less biased in their interpretations of NC. We present a corpus analysis with researcher annotations of several hundred tokens of these constructions in English treebank data. Initial results are consistent with previous conclusions based on controlled experimental data, suggesting that an attested tendency in linguistic comprehension is also reflected in naturalistic production. Our dataset can further be used to explore factors that modulate the disambiguation in context.
Based on results from a large-scale brain imaging study (https://doi.org/10.31234/osf.io/49uxp_v1), we found that the bilateral temporooccipital cortex is specifically involved in the encoding of aversive memories. Although these imaging findings are robust, they are correlational and do not imply causality. To investigate a causal relationship, we conducted a follow-up pilot TMS study (unpublished results) using a 30 Hz cTBS protocol. In that study, we observed a small effect size for aversive memory performance (number of freely recalled pictures) and a medium-to-large effect size (Cohen’s d = 0.77) for valence ratings of aversive stimuli: participants in the experimental group rated aversive pictures less negatively compared to those in the control group. Given the well-established link between emotional evaluation and memory performance, we hypothesised that the temporo-occipital cortex contributes to aversive memory performance through a mediating effect on the emotional perception of these stimuli. The primary objective of this study is to investigate the effect of bilateral temporo-occipital 30 Hz continuous theta burst stimulation (cTBS) on valence ratings of aversive pictures, in comparison to an active control condition (30 Hz cTBS applied to a bilateral region not involved in the processing of aversive stimuli). We hypothesise that aversive pictures will be rated as less emotionally aversive in the experimental condition compared to the control condition. Primary outcome will be the mean valence rating of aversive pictures as assessed by a visual analogue scale assessed directly after TMS.
This study investigates differences in artificial intelligence (AI) literacy and adoption between engineering students and faculty in a Middle Eastern higher-education institution. Parallel surveys were administered to undergraduate engineering students (N = 73) and faculty members (N = 20), each rating their familiarity with 20 AI tools covering learning, coding, productivity, and engineering applications. An AI Literacy Index was computed by assigning numerical values to familiarity ratings (A = 2, B = 1, C = 0) and normalizing the total to a 0–1 scale. Results from Welch’s t-test indicated that students demonstrated significantly higher literacy than faculty (0.454 vs. 0.356, p ≈ 0.042). Students also reported strong AI adoption for academic tasks (71.2%) and high perceived learning benefits (83.6%). Conversely, faculty expressed substantial concern about student over-reliance on AI (90%) while indicating readiness for professional development through AI training workshops (75%) and reporting assessment redesign efforts (75%). Overall, the findings highlight a meaningful literacy and perception gap with implications for engineering pedagogy, curriculum development, and assessment practices. Recommendations are provided to support the alignment of student and faculty AI competencies within engineering programs.
The article examines the queer annotated translation of the novel 39 rue de Berne (2013), written by Swiss Cameroonian author Max Lobe, aiming to spark discussions about the translation choices employed.The proposed translation is initially motivated by the representation of dissident sexuality in the work, bringing it closer to Brazilian LGBTQIA+ adult readers, as well as by adopting a questioning stance toward social and linguistic norms, a characteristic of queer.Drawing on passages from the central chapters of the novel, where issues of gender and sexuality are most evident, the research is imbued by theoretical contributions that outline translation strategies to Orlando
We present EmoWork, a multimodal, multi-label dataset designed to support emotion and stress detection in realistic interpersonal work settings. Interpersonal work—common in occupations such as customer service—often requires workers to regulate their emotional expressions in response to strong affective stimuli. These demands, shaped by organizational display rules, present a unique challenge for affective computing systems, particularly in scenarios where internal emotional states diverge from observable behaviors.Despite this, no public datasets exist that capture such dynamics of affect in naturalistic settings. To address this gap, we collected physiological, behavioral, and self-reported data from call center workers who engaged in role-play scenarios simulating customer service interactions with professional actors portraying dissatisfied customers. The dataset includes self-reported affective ratings, which are used as labels for classification, synchronized recordings from three wearable devices (i.e., Polar H10, Empatica E4, and Muse S), and features extracted from video and audio data. The EmoWork dataset advances affective computing by offering context-rich, multimodal data grounded in realistic interpersonal work scenarios.
Despite the growing use of NLP in second language (L2) research, model accuracy in L2 settings remains underexplored. This study addresses this gap by evaluating and fine-tuning a Korean language model to extract morphosyntactic features (i.e., morpheme tokenization/tagging and dependency parsing) from L2-Korean texts. We begin by evaluating a domain-general Korean language model on a gold-annotated L2-Korean treebank. We then fine-tune the model on L2-Korean data and quantify the resulting gains across diverse L1- and L2- datasets. Finally, we examine how model reliability varies with learner proficiency scores. Three key findings emerge: while the domain-general model excels at morpheme tokenization, it underperforms on morpheme tagging and dependency parsing; fine-tuning substantially improves adaptability to L2 morphosyntax; and proficiency has minimal effect on morpheme-level tasks but significantly affects dependency-parsing reliability. These results highlight the importance of incorporating L2 training data to improve morphosyntactic analysis in L2 settings and caution against uncritical reliance on automated dependency annotations, especially when performance varies across proficiency levels.
While most previous studies of evaluative conditioning have examined changes in emotional word meaning, few studies have examined the conditioning of gustatory word meaning. This study investigated whether the sweet taste meanings associated with real food names are linked with nonsense words. Participants completed the first evaluation phase, a conditioning phase, and the second evaluation phase in an experiment pairing Japanese nonsense words with Japanese sweet and non-sweet food names collected in a preceding survey. In both evaluation phases, they rated the sweet taste meaning, familiarity, valence, and arousal associated with nonsense words. The results revealed that nonsense words combined with sweet food names induced a sweeter taste meaning than did those combined with non-sweet food names after the conditioning phase. Despite controlling for familiarity, valence, and arousal across conditions, nonsense words paired with non-sweet food names were rated as more exciting than those paired with sweet food names. Familiarity and valence ratings were higher after the conditioning than before in both conditions. These results suggest that evaluative conditioning using sweet food names can promote sweet meanings and preferences of new food product names.
Horses are depended on as work animals by humans and are used in leisure and sport across the world, but the extent to which humans can recognise pain in horse faces is not known, which could impact their welfare. There are also significant gaps in our understanding of which psychological traits influence recognition of human facial expressions of pain. To address this, one hundred participants, with either some (N = 30) or no prior horse care experience (N = 70), rated thirty human and thirty horse faces for pain, arousal and valence and completed trait measures of empathy and social anxiety. Ten equine behaviour professionals also rated the horse faces as a baseline for assessing accuracy. Overall, accuracy of pain recognition was higher for human faces, but participants with horse experience were more accurate at pain recognition in horse faces, than those without, and years of horse experience predicted horse pain recognition accuracy. Social anxiety traits predicted accuracy of pain recognition in human but not horse faces, while also predicting subjective ratings of pain in horse but not human faces. Empathy and its cognitive and emotional components were not related to pain recognition accuracy or ratings of horse or human faces. Relationships between trait measures and arousal and valence ratings for both species are reported. This study is the first to report the human ability to read pain in horse faces and the factors which influence this and extends current knowledge on face processing in social anxiety.
Background. In today’s world, intercultural dialogue with other countries is actively developing and being realized through many forms of art. These include various literary works, multi-genre music, and a huge amount of audiovisual material, among which movies and TV series occupy an honorable place. Asian TV series (“doramas”) are one of the most popular in the world nowadays. Having started their journey in Korea, Japan, and China, they have spread throughout Asia, the United States, Europe, and even Africa over the past decade. Eastern cinema has not bypassed Ukraine either. Nowadays, many people in Ukraine enjoy watching Korean and Japanese films and TV series, learning about the mysterious culture of the East through them, and fan-sub groups are actively developing, translating and posting Asian films and TV series on thematic pages. In this regard, the issue of high-quality translation and adaptation of oriental films and TV series into Ukrainian is seen as an important aspect of further development of Oriental studies in Ukraine, popularization of the above-mentioned Far-Eastern cultures in Ukraine, and will also contribute to the deepening of intercultural exchange between our countries. Methods. The following methods were used in the course of the research: search, analytical, classification, and generalization. Results. The article examines the specifics of translating titles of films and TV series from Korean and Japanese on the basis of film titles selected from Ukrainian websites and other streaming platforms. The article discusses and traces the classification of different types of titles by structure, strategies and methods of their translation. The most common mistakes in the translation of titles and the reasons for their occurrence are analyzed. The author concludes that an accurate, complete translation of the title of a film product involves preserving all the functional and pragmatic features of the original in the language of the potential audience. However, more often than not, retaining an attractive title is a priority for distribution. Accordingly, the final title of a project does not always depend on the translator and can be reworked by distributors for the sake of marketing. Conclusions. It was found that the main approaches to translating film and TV series titles are the following strategies: descriptive translation, literal or direct translation, transcoding, transformation, genre translation, and semantic translation. The most common mistakes in the translation of film titles from Korean and Japanese into Ukrainian are due to the translator’s lack of background knowledge, ignoring allusions and other hidden elements in the original text, due to the inability to identify the stylistic features of the original title, violation of the norms of phonetics and readability, and mistakes when the title of a film or series does not correspond to its content.
The Philippines, having over 170 languages, faces language barriers and diminished language fluency due to heavy English influence and the effects of globalization. The language focuses on Cebuano, the second most spoken language mainly spoken in Central Visayas. Existing Cebuano parsers have been made through Recursive Descent Parsing (RDP) for a Cebuano Parse Tree in relaying grammar through a syntactic structure. However, there were identified limitations through low-level sentences, lacking grammar codes, low metric score, and stemmer evaluation errors. To address this gap, the researchers proposed a Snowball-based stemmer approach along with the Cocke-Kasami-Younger algorithm and additional grammar codes for better syntax tree optimization. Due to the lack of a dataset for Cebuano sentences, the Cebuano Corpus is extensively created by the researchers via web scraping and data augmentation techniques such as language dialect filtration and sentence range variation. The dataset comprises at least 500 collected phrases to be annotated by the linguist experts for the Gold Standard Parse Treebank (GSPT). The GSPT will serve as the parser's benchmark through syntax constituents, known as the PARSEVAL metric system. The improved parser attained an accuracy of 82.14%, suggesting a high rate of correct Cebuano constituents from a Cebuano sentence. The findings suggest that the developed system is more effective than its predecessor. By leveraging advanced Snowball-based algorithms and updated Grammar Codes, the study demonstrates the potential of parsers in preserving cultural language.
The linguistic features of the Uzbek language - complex agglutinative morphology, free word order, and limited resources - necessitate a specialized approach and thorough research in the application of morphological and syntactic methods. Within the framework of the study, morphological analysis methods and syntactic analysis methods are reviewed based on scientific sources. Each section presents the existing advantages and disadvantages, experience of their use in the Uzbek language, as well as a comparative analysis with foreign languages. Rule-based methods, statistical models (HMM, CRF, etc.), Neural network-based approaches (BiLSTM-CRF, seq2seq) of morphological analysis in the Uzbek language are discussed, and the results are given in examples and percentages. It is shown that syntactic parsing is implemented using dependency and constituency parsing analysis methods. The issue of building a UD treebank for the Uzbek language with SOV order is considered. The impact of complex morphological structure and free word order in sentences on the construction of parsers is highlighted. As a result of the studied approaches, the issue of building hybrid parsers, integrating them with morphological analysis and assigning grammatical categories of words to the parser is raised. Also, the development of neural constituency parsers based on neural networks and the effectiveness of the results obtained from them are analyzed.
The article is devoted to the study of the current topic of modern communication– virtual language personality, and the description of lexical and semantic features of virtual language personality in Internet communication. The issue of dynamic changes in the Kazakh vocabulary introduced through Internet communication has become especially relevant in recent years. The research work began with a review of the research of domestic and foreign scientists, the features and main trends were described. During the review of scientific papers analyzing the discourse of the social network and its text, the fundamental concepts were identified and the necessary conclusions were drawn, which guided us. The purpose of the study Is to consider the set of lexical and semantic properties of social network usages. The article notes that the main feature of communication in Kazakh-speaking social networks is the replacement of a real participant by a participant in virtual communication. We have done analysis of the structure, lexical and grammatical characteristics of Kazakh texts collected from social networks. The features of the texts of the discourse of social networks were described. It is mentioned that texts in social networks are not a source of knowledge, they are news or entertainment information, and accordingly they should have brief but rich information, the information in the text is transmitted not only by writing letters, but also by pictures, audio and video recordings. Examples from Facebook and WhatsApp networks were given as research materials. The study revealed the use of Facebook language terms in Kazakh texts, an increase in the activity of foreign words, a lexical-syntactic combination of foreign and native words, a combination of words without a lexical combination, etc. As a result, on the basis of the collected materials, grammatical (morphological, syntactic) features of Kazakh texts in the social network were divided into groups and a linguistic analysis was carried out. In modern public consciousness, linguistic research has become relevant, defining the linguistic image of a social network as a means of implementing linguistic communication. In this regard, it is very important to improve the language skills of network users.
Large language models (LLMs) have gained a lot of attention and achievements recently because of their significant comprehension and generative abilities. However, the large-scale parameters of LLMs require considerable computational resources in the training and inference process, which restricts their wide application. To overcome this challenge, we propose an efficient mixed precision weight quantization (EMWQ) method for LLMs in this article. Specifically, we introduce a new outlier detection method by analyzing the weight distribution instead of the conventional weight magnitude. Then, we propose a dual-quantization strategy that quantizes both the outlier critical columns and the residual matrices with different precision. Besides, we introduce two effective EMWQ-based application frameworks, the EMWQ-R and EMWQ-O in our study. Comprehensive experiments are conducted on the Penn Treebank (PTB), C4, ARC-Easy datasets, and MMLU benchmark across various tasks. The comparison results demonstrate that the proposed EMWQ achieves state-of-the-art performance in mixed precision quantization and further reduces computational memory cost. Besides, it has higher generalizability compared with conventional methods.
This is an accepted article with a DOI pre-assigned that is not yet published.The purpose of this paper is to test the correlation between V2 word order and its presumed micro-cues. Micro-cues are structural manifestations that have been suggested to associate to a syntactic phenomenon, making its organization and acquisition accessible to users and learners. Three micro-cues of declining Medieval French V2 identified in the literature are tested for their relation using a treebank method. Statistical analyses confirm the expected correlation. The results lend plausibility to a micro-cue model for the acquisition and structuration of syntactic phenomena.
This article explores the role of five expressive punctuation marks – multiple exclamation marks (!!!), exclamation mark (!), full stop (.), ellipsis (...), and null punctuation (ø) – as cues to writer attitudes in CMC. Specifically, it investigates how the underlying meaning of expressive punctuation influences the perceived emotional valence of discourse referents within exclamative constructions. In a 1x5 between-subjects repeated measures design, valence ratings were collected for 120 discourse referents embedded in exclamative constructions manipulated by message finale punctuation mark (e. g., What a view!!!/!/./…/ø) on a 1 (negative) to 9 (positive) scale. For inherently positive discourse referents, a clear positivity hierarchy in the overall valence of embedded discourse referents emerges, indicating a differential influence of punctuation on perceived valence: multiple exclamation marks > exclamation marks > null punctuation > full stop > ellipsis. For inherently negative discourse referents, the differences between the conditions are less distinct. Notably, only multiple exclamation marks yield a significantly lower valence rating within that range of values. While the findings for inherently positive referents align with prior assumptions on the expressive meaning of different punctuation marks in CMC, the observed pattern for inherently negative referents cannot be readily explained by existing literature on expressive punctuation in CMC.
The current study examined how adolescents respond to emotionally arousing images that are considered age-appropriate, such as sports, food, and threatening animals. Prior research in adults has shown a mismatch between subjective arousal ratings and the late positive potential (LPP) for pleasant images, and we aimed to test whether this discrepancy is also present in adolescents. The results showed a larger LPP in response to unpleasant, compared to pleasant, contents. However, comparisons across the different emotional contents showed that the concordance between subjective and neural engagement varies depending on the emotional contents of the scenes. Specifically, images of threatening animals were rated high in emotional arousal and prompted the largest LPP in adolescents, while images of sad people and images of mundane activities were rated lower in emotional arousal and prompted the smallest LPP responses. However, adolescents showed significant divergence in their responses to pleasant images depicting food and sport: despite being rated high in emotional arousal, these images elicited relatively small LPPs. These results highlight the challenges of selecting emotional pictures to assess neuroaffective responses to pleasant and unpleasant pictures in adolescents. To overcome these challenges, future studies may adopt experimental paradigms that will allow researchers to measure neuroaffective responses not just in free-viewing contexts but also during anticipation and reception of actual rewards (and punishments).
This study aims to examine Translation Management Systems (TMS) and their role in knowledge generation by exploring the mechanisms for organizing and automating the process of machine-based knowledge generation on the Phrase platform, as well as evaluating the effectiveness of artificial intelligence integration in this context. The study also seeks to assess the reliability and security of data on the Phrase platform and to identify the core values and principles it upholds in knowledge generation and the management of scientific terminology. The researchers employed a content analysis methodology to investigate the platform’s role in knowledge generation. The study concluded with several findings, most notably that Phrase supports the four stages of knowledge creation as outlined in the SECI model. The Socialization stage was evident in the exchange of experiences among translators through forums and workshops within the platform. Externalization appeared through the documentation of translators’ experiences in the form of articles, guides, and manuals, as well as through the use of artificial intelligence in translation and analysis. Combination was observed through the integration of linguistic databases, terminology resources, and translation tools with various sources to develop more accurate translation models, alongside the use of machine learning techniques to build an interactive translation database. Finally, Internalization was reflected in the training of translators through educational courses based on stored explicit knowledge, aimed at enhancing performance. The researchers recommend that institutions involved in knowledge generation should leverage translation management systems such as Phrase, capitalizing on its data reliability, security features, and its foundational values and principles in knowledge generation and terminology management. Additionally, they recommend the inclusion of dedicated courses on translation management systems within university-level translation programs.
Number of images in each category of the dataset along with their average valence and arousal ratings provided by human participants. The table also illustrates the distribution of human ratings for valence and arousal using the Likert scale.
Uncertainty of scientific findings are typically reported through statistical metrics such as $p$-values, confidence intervals, etc. The magnitude of this objective uncertainty is reflected in the language used by the authors to report their findings primarily through expressions carrying uncertainty-inducing terms or phrases. This language uncertainty is a subjective concept and is highly dependent on the writing style of the authors. There is evidence that such subjective uncertainty influences the impact of science on public audience. In this work, we turned our focus to scientists themselves, and measured/analyzed the subjective uncertainty and its impact within scientific communities across different disciplines. We showed that the level of this type of uncertainty varies significantly across different fields, years of publication and geographical locations. We also studied the correlation between subjective uncertainty and several bibliographical metrics, such as number/gender of authors, centrality of the field's community, citation count, etc. The underlying patterns identified in this work are useful in identification and documentation of linguistic norms in scientific communication in different communities/societies.