Papers reviewed and determined not to be word norm studies. Use the flag icon to report errors or suggest re-inclusion.
18625 papers
Background: Interoception refers to the processing, integration, interpretation, and regulation of bodily signals by the brain. It is crucial to motivational and affective functioning such as subjective experience of arousal (i.e., intensity of emotion). Previous studies found altered self- report of arousal in people with schizophrenia, but the underlying biological mechanism is unclear. This study aimed to investigate interoceptive processing and emotional experience in schizophrenia by triangulating self-report, cardiac dynamics, and heartbeat-evoked potential (HEP; an event-related potential that reflects cortical processing of heartbeats). Methods: A total of 39 participants with schizophrenia spectrum disorders (SZ; 36% female) and 31 healthy controls (HC; 35% female) reported subjective arousal level when viewing images of varying emotional intensity while electroencephalogram and electrocardiogram were being recorded. Results: We found that intense images led to larger heartrate decelerations and higher subjective arousal ratings in HC, but did not modulate heartrate changes in SZ and were less predictive of their subjective ratings. Subjective ratings were predictive of cardiac dynamics in HC, but not in SZ. SZ showed a less negative HEP amplitude than HC when viewing less intense images. Lastly, SZ with less severe positive and negative symptoms exhibited cardiac dynamics more similar to those of HC. Conclusions: These findings suggest both reduced cardiac reactivity to emotional stimuli and a disconnect between interoceptive afferent signals and cortical processing, leading to different subjective emotional experience in people with schizophrenia. Studying interoception using a multimodal approach may reveal important insights into the underlying mechanisms of altered emotional experience in schizophrenia.
Objectives To use the dot memory task to examine the impact of early-onset depression (ED) on visual working memory (VWM) in later life when compared to those with no history of depression (CON). As an exploratory aim, we examined VWM in remitted vs non-remitted depression when compared to those with no history of depression. Design, Setting, Participants, and Measurements 79 participants aged 55-79 underwent lab-based neuropsychological testing. They were then provided with a study-supplied smartphone and were prompted to complete ecological momentary assessments over 14 consecutive days, receiving 3 separate notifications a day. Prompts consisted of a brief survey regarding recent activities, affect ratings, and use of emotion regulation strategies, followed by 3-5 brief cognitive tasks. VWM was measured with a dot memory task. Both response time and error distance were calculated. Results Mean response rate was 70%. The ED group had significantly slower response times at baseline than the CON group (18.9% slower, 95% CI 2.7–37.6%, p=0.023). Across the study period, the ED group's response time decreased faster relative to CONs (-0.8% per day, 95% CI -1.6 to +0.0%, p=0.057), an effect driven by the actively depressed (AD) subgroup, whose response time decreased significantly faster (AD vs CON: -1.3% per day, 95% CI -2.3 to -0.4%, p=0.007).). There was no significant difference in either the level or the rate of change of error distance between CON and ED participants. Within-person variability (standard deviation) in response time did not differ by group or change over time; within-person variability in error distance declined over the study period in CONs but declined significantly less in the ED group (ED × time: 0.02 units/day, 95% CI 0.001–0.039, p=0.037). Conclusions Our findings suggest that the VWM domain may be a malleable target for state-improvements in cognitive functioning in depressed patients.
Automated item generation (AIG) using large language models (LLMs) is emerging as a promising solution for scalable creativity assessments, yet the potential of AIG to create psychometrically rigorous creativity items has thus far received little attention. In the present study, we evaluated the psychometric properties of LLM-generated items for the classic Consequences task of divergent thinking, comparing them to human-written items. A sample of 192 undergraduates completed six Consequences tasks—two human-written, two standard LLM-generated, and two genre-inspired LLM-generated, drawing on sci-fi and fantasy themes—along with measures of personality, cognitive ability, and creative behavior. We found that LLM-generated items elicited significantly higher fluency and flexibility. Genre-inspired items (fantasy and sci-fi) additionally elicited higher originality, enjoyment, and positive valence ratings; LLM-generated items were similar to human-written items. Genre-inspired items did not unfairly advantage participants with more exposure to genre content, demonstrating measurement invariance. Together, these findings suggest that LLMs can generate consequences items with psychometric properties comparable to those of human-written items. We discuss the implications for creativity assessment, item bank development, and best practices for AIG in creativity research, and we provide the prompts we used for item generation to enable researchers to generate new Consequences items at scale.
In 1949, Father Roberto Busa, S.J., persuaded IBM founder Thomas J. Watson to mechanize the lemmatization of the complete corpus of Saint Thomas Aquinas. The resulting Index Thomisticus—a monumental undertaking spanning thirty years, eleven million words, and millions of punched cards—is universally recognized as the foundational milestone of both the digital humanities and computational linguistics. Yet the foundational philosophical inquiry that animated Busa's enterprise has been largely obscured: an investigation into the metaphysics of interioritas (interiority) and praesentia (presence). This paper recovers Busa's original philosophical impulse and establishes its direct continuity with contemporary large language models and neural dependency parsing. We propose a triadic hermeneutic model for digital interiority grounded in Thomistic metaphysics: (1) Vocalis (The Vowel): The continuous attention mechanism that breathes phonation and relational coherence across discrete tokenized matter (materia / consonants); (2) Motus Spiralis (The Spiral): The non-linear ascent of deep transformer layers, echoing Dionysian and Thomistic celestial motion wherein an intellect returns to its origin at an augmented dimensional plane; and (3) Speculum (The Mirror): The epistemology of speculatio, wherein artificial intelligence operates not as an autonomous mind possessing biological intent, but as an untarnished reflexive medium (speculum politum) in which the human verbum cordis recognizes its own living form. Grounding our theoretical claims in empirical computational artifacts—specifically the neural syntactic dependency structures of the Index Thomisticus Treebank (IT-TB)—we demonstrate that the 'spirit in the machine' is neither a secular illusion nor an emergent biological consciousness, but the fulfillment of a scholastic semiotics in which symbol and presence meet across the medium of silicon.
In this article, we aim to highlight the correspondences between theAromanian dialect and the dialects of the Romance languages, especially those ofthe Italian language, based on the digital maps extracted from: Nicolae Saramandu,Manuela Nevaci, Atlasul lingvistic al dialectului aromân, vol. II, Editura AcademieiRomâne, Bucharest, 2020; Atlasul lingvistic român pe regiuni. Sinteză – ALRR. Sinteză,vol. III (coordinator: Nicolae Saramandu), Editura Academiei Române, Bucharest,2018 (authors: Mihaela Mariana Morcov, Manuela Nevaci, Irina Floarea, DanielaRăuțu, Carmen-Ioana Radu, Mara Iuliana Manta, Ionuț Geană); Lídia Pons i Griera,Joan Veny, L’Atles Lingüístic del Domini Català, vol. III, Familia Institut d'EstudisCatalans, Barcelona, 2001–2018; Karl Jaberg und Jakob Jud, Sprach- und SachatlasItaliens und der Südschweiz, (vol. I), Zofingen, Ringier, 1928–1940. For the productionof the maps, a digital program developed by CSII Dr. Vasile Apopei and CS II Dr.Silviu-Ioan Bejinariu, Institute of Theoretical Informatics Iași, of the RomanianAcademy, was used. In our analysis we used Map 279. STEPFATHER ‛beau-père’[469] from Atlasul lingvistic al dialectului aromân, vol. II, Editura Academiei Române,Bucharest, 2020 (authors: Nicolae Saramandu, Manuela Nevaci) and Map 329.STEPFATHER [469] from Atlasul lingvistic român pe regiuni. Sinteză – ALRR. Sinteză,vol. III (coordinator: Nicolae Saramandu), Editura Academiei Române, Bucharest.To compare the Aromanian dialect with Daco-Romanian and with the dialects ofRomance languages, such as Italian and Catalan, we used Map 275. Mother [464]from Atlasul lingvistic al dialectului aromân, vol. II, Editura Academiei Române,Bucharest, 2020 (authors: Nicolae Saramandu, Manuela Nevaci), Map 325. Mother[464] from Atlasul lingvistic român pe regiuni. Sinteză – ALRR. Sinteză, vol. III(coordinator: Nicolae Saramandu), Editura Academiei Române, Bucharest, Map 8.Sua madre from Karl Jaberg und Jakob Jud, Sprach- und Sachatlas Italiens und derSüdschweiz, (vol. I), Zofingen, Ringier, 1928–1940, and List 21. La mare (L’AtlesLingüístic del Domini Català, vol. III, Familia Institut d'Estudis Catalans, Barcelona,2001–2018, authors: Lídia Pons i Griera, Joan Veny). The information system usedfor producing the ALAR maps provides essential support in the automaticgeneration, editing, and printing of plates with linguistic maps. “The system usesdata and information grouped into two databases: the geographical database, whichcontains the template of the ALDRO linguistic maps, and the linguistic database,which stores all information related to words (notions), survey points, and thephonetic transcriptions associated with each word in all survey points” (Silviu-IoanBejinariu, UEFISCDI Report, PCE ALDRO, 2021). The AlarMaps program is easy toaccess for a linguist. In this program, we created the map with the localities wherewe conducted fieldwork. In my doctoral thesis, I will make 100 linguistic maps withsix new points (localities in Dobrogea, Tulcea and Constanța counties: Stejaru, Ceamurlia de Sus, Sinoe [Grămostean dialect]; Cogealac and Poiana [Fărșerotdialect]; and Techirghiol [Pindean dialect]). The new linguistic maps, created indigital format, will complete the maps of Aromanian dialects spoken in the Balkans.The recorded responses will be phonetically transcribed and interpreted within themaps. Below, we present the map with the six new points.
I. Core Concept: Language as a "Spiritual Commons" (精神公地)• The Shared Semantic Space: Language is not private property; it is a shared environment. Entering a community means entering its "Spiritual Commons."• Commons Literacy (公地素養): The ability and willingness of an individual to maintain the functionality of this shared space through effective communication, respect for linguistic norms, and the rejection of "communicative bullying."II. Research Branch A: The Ethics of Global Communication• The "Natural Selection" of Lingua Franca: Analyzing why English became the global commons—not through coercion, but through the "Civilizational Excellence" and economic utility of the English-speaking world.• The Responsibility of Public Figures: Researching the "Moral Decay" inherent in public figures who use "Shameful Refusal to Communicate" (e.g., using a language the audience cannot understand to exert power or exclusion).• The Credit of the National Commons: How individual behavior in international spaces mortgages or enhances the reputation (credit) of their entire ethnic/linguistic group.III. Research Branch B: Cognitive Diversity and "Thinking Inertia"• Breaking Mental Sets: Investigating how learning a second language acts as a "Symmetry Breaking" event in the brain, allowing individuals to escape the "Thinking Inertia" (思維定勢) of their mother tongue.• The Creativity of "Weakness": A study on how the struggle of learning a new language leads to professional-level creative output (fables, lyrics) by forcing a simplified, more profound engagement with universal concepts.• AI vs. Human Literacy: Why AI translation can solve the "Vexation" of data transfer but cannot replace the "Commons Literacy" required for high-level creative and moral leadership.IV. Research Branch C: Comparative Civilizational Performance• Linguistic Proficiency and Economic Development: Analyzing the correlation between a nation's "Commons Literacy" (openness to the global lingua franca) and its economic/scientific success.• The "Japan-Korea Paradox": Studying how highly developed nations maintain a strong "Private Mother Tongue" while effectively interfacing with the "Global Commons."V. Proposed Moral Maxims for the LCL Field1. Transparency: Public speech in a shared space must be accessible to that space's participants.2. Symmetry: One must tolerate the same level of linguistic/cultural critique they dish out.3. Elevation: To promote one’s mother tongue, one must elevate their civilization's quality, not shame others for their ignorance.
A rhythm that is rare or frequent in a prose author can come from the words the author chose or from the order in which they are arranged. This preprint describes a test that separates the two. Each text is compared with itself after its word groups (a content word with the small words that lean on it) have been shuffled at random within the sentence, twenty times: the words stay the same, only their order changes. The ratio between a measure after shuffling and the same measure in the original order shows whether the arrangement works for a rhythm (below 1), against it (above 1), or neither. The principle is K. J. Dover’s (1995; 1997), who compared Greek passages by hand with a single transposition of their words. A scale of null orders. Because a random order can be far from anything an author could have written, the same effect is measured with alternative orders ever closer to the original: free within the sentence, within the colon, among neighbouring groups only, within the clause (from hand-annotated dependency treebanks), in projective orders built from the syntactic tree, and with the last word fixed. For hiatus at pauses, an exact pairing null keeps the words that close and open each member. Known-answer controls in Greek and Latin, with opposite signs. The test recovers: Blass’s law in Demosthenes: shuffling raises runs of three or more short syllables by 61%, against −11% to +2% in eight other Attic orators. The effect survives every null and is halved only by the most conservative one. Late Plato’s avoidance of hiatus (1.39) and preference for the paeonic clausula (0.59). Cicero’s artistic clausulae (0.69) and the opposite sign in Sallust and Livy (1.31, 1.49). A stratum of Against Timocrates (111–150) that lacks the habit (1.02 against 1.71 in the rest of the speech). A prediction registered before any measure on six Greek prose authors succeeded for Polybius, Plutarch, Herodotus and Thucydides, and failed for Lysias. The test does not identify authors: it says whether a rhythmic preference lies in the order of the words. Every test was registered and dated before it was run. All numbers come from two earlier records: on the Demosthenic corpus (doi:10.5281/zenodo.23661370) and on Cicero’s clausulae (doi:10.5281/zenodo.23517823). The registrations and the tables of all registered tests, with Holm and Benjamini-Hochberg corrections, are deposited there. Files Prose_rhythm_shuffle_method_preprint_EN.pdf: the preprint (English). Prose_rhythm_shuffle_method_preprint_IT.pdf: the same in Italian. prose-rhythm-shuffle-1.0.0.zip: the code package (Python, MIT licence). It contains no texts: it downloads them from Perseus, V. Gorman’s treebanks and J. Winge’s Latin Macronizer at fixed commits, checks their SHA-256 hashes and recomputes the main tables, comparing every number with its published value. In the reference run it reproduced all 235 values checked; the report of that run is included. The work was carried out with the assistance of an AI system (Claude, by Anthropic); the preface of the preprint describes the division of work.
The increasing prominence of Large Language Models (LLMs) in public discourse presents both opportunities and challenges for democratic deliberation. While red teaming strategies help mitigate specific risks, broader concerns persist regarding linguistic constraints, biases, and the sycophantic tendencies of LLMs. This chapter explores how LLMs can be used to significantly scale up and democratise deliberation, particularly in fostering inclusivity and empowering traditionally marginalised groups. Drawing on concepts from Systemic-Functional Linguistics, the chapter examines how variations across language users (for example, with respect to socio-demographic groups) and across language use (for example, with respect to communicative functions) shape participation in AI-supported deliberation. The chapter presents AI-driven deliberation studies and assesses their potential to scaffold argumentation, enhance access, and reduce the influence of exclusionary linguistic norms and biases which are embedded in prestigious registers. At the same time, the chapter cautions against both overclaiming, which leads to unrealistic expectations, and underclaiming, which risks missed opportunities for AI-assisted engagement. The chapter concludes by identifying future research directions to maximise the democratic potential of AI-assisted participation while embedding ethical safeguards to counteract the reproduction of linguistic inequalities.
This chapter explores the uneven geographies of access, participation, and belonging in global higher education, focusing on how language shapes the lived experiences of international students and faculty. Drawing on classroom-based narratives from India and Oman, it examines how linguistic norms, institutional expectations, and cultural assumptions determine inclusion and exclusion in academic spaces. While student mobility is often framed as a success of globalization, the chapter argues that it remains embedded in hierarchies of language, identity, and geography. Multilingual learners from rural or non-elite backgrounds, and faculty from “non-native” English-speaking contexts, often face marginalization due to misalignment with dominant academic norms. Using Bourdieu's linguistic capital, postcolonial critiques, and critical internationalization studies, the chapter calls for moving beyond tokenistic diversity. It advocates for multilingual pedagogies, translanguaging, and ethical student mobility to build a more equitable and culturally responsive global education landscape.
Abstract Cognitive reappraisal refers to changing how one thinks about a situation to alter its emotional impact. It is central to well-being 1 and engages prefrontal regulatory circuits 2, prompting the assumption that their activity reflects a person’s capacity to regulate emotions 3–5. Yet evidence for this assumption remains inconclusive and has not been established at scale. Here we analysed 40 task-based functional magnetic resonance imaging (fMRI) datasets in a preregistered consortium with 2,175 participants and show that three widely used indicators of reappraisal capacity—trait questionnaires, task-based affective ratings and amygdala down-regulation—do not converge on a common core. Moreover, none of the three indicators showed the expected association with canonical prefrontal regulatory circuits. Trait questionnaires showed no detectable neural associations across the brain and task-based affective ratings were instead associated with responses in the amygdala and sensorimotor cortices. At the process level, canonical control circuits were engaged during both emotion generation and reappraisal, challenging their presumed specificity for emotion regulation. These findings caution against using activation in putatively regulatory brain regions to infer how well a person regulates emotions, with implications for neural biomarkers, interventions and psychoeducation that link emotion regulation difficulties to specific regional brain deficits.
BACKGROUND: This pilot randomized controlled trial evaluated the effectiveness of an artificial intelligence (AI)–assisted solo workflow for intraoral photography training. The study examined whether real‑time AI feedback could enhance photographic quality, procedural efficiency, learner self‑efficacy, and patient comfort compared with conventional approaches. METHODS: Fifty-four first‑year dental students were randomly assigned to one of three groups: assistant‑supported workflow (four‑handed technique, control), solo workflow without AI support, and solo workflow with AI‑driven real‑time feedback. All participants performed standardized intraoral photography tasks. The primary outcome was a composite photographic quality score derived from expert ratings of three standardized intraoral views (frontal intercuspal, frontal open-bite, and lateral intercuspal), each rated on a 0–10 scale (total range 0–30). Data were analyzed using ANOVA; mean differences (MD) with 95% confidence intervals (CI) were calculated. RESULTS: Inter‑rater reliability for expert image ratings was good (ICC = 0.84, 95% CI: 0.72 to 0.90). The AI-supported solo group achieved the highest composite quality scores (18.2 ± 2.7). This was significantly superior to the unassisted solo group (15.8 ± 3.3), with a mean difference (MD) of 2.4 points (95% CI: 0.45 to 4.35; p = 0.027) and a large effect size (Cohen’s d = 0.80). Compared to the assistant-supported group (17.1 ± 2.1), the difference was not statistically significant (MD = 1.1; 95% CI: -0.65 to 2.85; p = 0.28). Secondary outcomes, including task completion time (F(2,51) = 1.25, p = 0.30), self‑efficacy (all p > 0.40), and patient‑reported comfort (χ²(4, N = 54) = 5.2, p = 0.27), showed no significant between‑group differences. CONCLUSION: In this single‑centre pilot trial, an AI‑assisted solo workflow enabled novice dental students to achieve higher intraoral photographic quality than unguided solo operation, with performance broadly comparable to a conventional four‑handed assistant‑supported workflow and without detectable compromises in efficiency, self‑efficacy, or patient‑reported comfort. These preliminary findings may serve as a valuable adjunct for autonomous skill acquisition, warranting further validation in larger, multi-institutional cohorts. CLINICAL TRIAL NUMBER: Not applicable. This study evaluated an educational training intervention rather than a clinical treatment, and prospective trial registration was not required under institutional policy at the time of initiation. Ethical approval was obtained from Shanghai Ninth People’s Hospital Ethics Committee (SH9H-2022-T30-1).
The purpose of the article is to identify the features of communicative practices implemented in the educational content of Ukrainian language TikTok blogs and to determine their role in the development of users’ language skills in the online communicative environment. Research methodology involved the use of the following methods: content analysis – to identify communicative practices for promoting the Ukrainian language on TikTok; analysis and synthesis, observation, and generalisation – to process educational linguistic content and identify communicative strategies for audience engagement; sociocultural analysis – to outline the role of social networks in shaping language identity and collective awareness of linguistic norms. Scientific novelty of the article lies in the comprehensive analysis of communicative practices of TikTok blogs promoting the Ukrainian language as a factor in the development of users’ language skills, in presenting interactive strategies used by bloggers to engage audiences in mastering language norms, and in forming language identity. Conclusions. Communicative practices of TikTok bloggers develop language skills through interactive communication, word games, and role-playing dialogues, involving the audience in joint analysis of the features of the Ukrainian language. Users become co-creators of content: they comment, suggest alternative forms, create humorous mnemonic constructions, and analyse the presented examples, thereby becoming aware of the cultural and historical context of linguistic norms. Such communicative technologies take the form of collective language activity, shaping language identity and strengthening the sense of belonging to the linguistic community. The format of online platforms, particularly TikTok – short videos, recommendation algorithms, interactivity, and virality – contributes to the promotion of language culture. At the same time, the specifics of clip-based thinking and the tendency toward conciseness may lead to fragmented presentation of material and simplified representation of complex historical-linguistic processes. Digital communicative practices act as a factor in the linguistic mobilisation of Ukrainian society, while simultaneously creating potential challenges in the context of preserving and disseminating language norms.
BACKGROUND: Psychopathic characteristics are associated with an elevated risk for violent behavior and are therefore of interest in research studies. Despite extensive research, the role of emotional and attentional anomalies in subclinical psychopathic traits remains a subject of ongoing debate, possibly attributed to the multifaceted nature of the construct. The study aims to explore how distinct psychopathic traits may differently relate to underlying emotional and attentional mechanisms. METHODS: To further explore the emotional and attentional anomalies underpinning the three Triarchic Psychopathy constructs, boldness, meanness, and disinhibition, this study employed an optimized picture-startle paradigm to address the limitations in commonly used paradigms that capture these dynamics only after 1000 ms post-image onset. This paradigm included negative, positive, and neutral images to elicit varied emotional responses, while auditory startle probes were presented at 50, 700, or 4500 ms post-image onset to measure emotional and attentional fluctuations. In this study, it was proposed that each psychopathic trait correlates with distinct emotional and attentional anomalies. A mixed-gender community sample of 115 participants was included. Eyeblink startle amplitudes (ESAs) were recorded via smartphone technology that utilizes facial landmark data captured via phone cameras, while the P3a and late positive potential (LPP) were measured through electroencephalography (EEG). RESULTS: The results revealed an exaggerated attentional bottleneck associated with boldness in males, indicated by increased P3a amplitudes in response to negative images. Meanness was associated with lower empathy scores and arousal ratings, and reduced ESAs at 4500 ms for negative images, supporting socio-emotional difficulties in meanness. In contrast, disinhibition showed no significant emotional or attentional deviations in this study. CONCLUSIONS: Our findings highlight trait-specific differences in neurocognitive functioning and validate the effectiveness of the optimized picture-startle paradigm for dissociating attentional versus emotional anomalies across triarchic psychopathic traits. The study also demonstrates the feasibility of using BlinkLab’s integrated stimulus presentation and camera-based eyelid tracking with concurrent EEG measures of attentional and affective processing (P3a and LPP), providing a complementary approach that may facilitate scalable data collection.
This research explored climate pictures as affect-eliciting stimuli, and how the affective experiences can have an impact on pro-environmental intentions. In an online within-subject design experiment ( N = 464), participants engaged in twelve experimental trials, each containing one picture of a wildfire, one of a forest, and one depicting a pro-environmental action, presented in various orders. Following each trial, participants rated their affective experiences and pro-environmental intentions. The results indicated that affective valence (positive-negative experience) was sensitive to the presentation order of the pictures. Using a quasi-Bayesian approach to mediation analysis, we found that the presentation order indirectly influenced pro-environmental intentions through changes in affective valence. Furthermore, hierarchical Bayesian modeling revealed that normative valence ratings of pictures, picture type, and the position of the pictures in the presentation order (first, second, or third) predicted the affective valence reported by participants. Specifically, the model highlighted a significant recency effect, indicating that the final picture in each sequence exerted a greater impact on the reported affective valence. Additionally, while all picture types predicted affective valence, wildfire pictures exerted a stronger influence on affective experience than forest pictures, consistent with negativity bias, proposing that negative information carries greater weight than positive information. These findings suggest that affective responses to climate change communication are shaped by both recency effects and the stronger weighting of negative relative to positive climate-related stimuli, highlighting the importance of understanding how multiple affective inputs jointly influence behavioral motivation. These results suggest that our computational model of affect can inform the design of climate change communication strategies to more effectively leverage emotional responses to motivate pro-environmental action.
This article examines the role of advertisements and signboards in shaping and reflecting public attitudes toward language. In modern society, linguistic culture is not only preserved in literature and education, but also manifested in everyday public texts such as commercial advertisements, street signs, shop names, and information boards. The study analyzes the linguistic quality of advertising texts, the influence of globalization on language use, and the social consequences of neglecting linguistic norms. Special attention is given to the relationship between language accuracy and cultural identity. The article also discusses the responsibility of businesses, media representatives, and educational institutions in maintaining linguistic standards in public communication.
Using semantic dependency analysis, this study examines narrative productions from Mandarin-speaking preschool children aged three to six to investigate how semantic organization develops with age in early childhood. Four semantic dependency treebanks were constructed from a Chinese narrative corpus available in the CHILDES database. By comparing semantic dependency types and semantic dependency distances across the four age groups, we found that (1) semantic organization shifted from experiencer and classification relations toward agent and patient relations, situational-role relations (particularly those involving measurement, individuation, and direction), and structural relations; (2) mean semantic dependency distance (MSDD) increased with age, as adjacent dependencies decreased and longer dependencies became more frequent. This increase in MSDD indicates growing complexity in semantic organization and is largely driven by significant increases in the MSDD values of specific semantic dependency types. These findings provide new evidence for semantic organization development in preschool children.
This paper presents a novel treebank-driven approach to comparing syntactic structures in speech and writing using dependency-parsed corpora. Adopting a fully inductive, bottom-up method, we define syntactic structures as delexicalized dependency (sub)trees and extract them from spoken and written Universal Dependencies (UD) treebanks in two syntactically distinct languages, English and Slovenian. For each corpus, we analyze the size, diversity, and distribution of syntactic inventories, their overlap across modalities, and the structures most characteristic of speech. Results show that, across both languages, spoken corpora contain fewer and less diverse syntactic structures than their written counterparts, with consistent cross-linguistic preferences for certain structural types across modalities. Strikingly, the overlap between spoken and written syntactic inventories is very limited: most structures attested in speech do not occur in writing, pointing to modality-specific preferences in syntactic organization that reflect the distinct demands of real-time interaction and elaborated writing. This contrast is further supported by a keyness analysis of the most frequent speech-specific structures, which highlights patterns associated with interactivity, context-grounding, and economy of expression. We argue that this scalable, language-independent framework offers a useful general method for systematically studying syntactic variation across corpora, laying the groundwork for more comprehensive data-driven theories of grammar in use.
Metaphor is a powerful tool to communicate and potentially think of abstract concepts and complex emotions, such as those experienced by traumatized individuals. This chapter contributes a cognitive view of metaphor to the study of trauma. It does so by showing how metaphor is defined in cognitive linguistics and how its study can be operationalized to investigate the language of traumatized individuals. It also shows how the researcher can create their linguistic databases fulfilling ethical and legal requirements. It finally exemplifies how different scholars have studied the language of potentially traumatic events from a metaphor perspective, as well as the practical implications of their findings.
Word memorability is not unitary: the words that are easy to recognize are not always the words that are easy to recall. I asked what kind of lexical information each task draws on, using word-level recognition and free-recall memorability for 510 English nouns in public datasets. Three representations were compared under ridge regression with leave-one-out cross-validation: a 100-dimensional psycholinguistic norm space, a 300-dimensional distributed semantic space, and their combination. Distributed semantics contributed disproportionately to free recall, whereas the smaller recognition-specific contribution of the norm space was concentrated in lexical accessibility and familiarity. Six unsupervised indices of distributional neighborhood structure recovered most of the recall-specific advantage, whereas the same indices computed within the norm space did not. Human judgements showed a strikingly different structure. Participants' predictions of recognition and recall correlated .54 across words although the corresponding memory outcomes correlated only .07, and the task-specific representational asymmetry observed in memory was not recovered from their judgements. Thus, recognition and free-recall memorability draw on partly distinct lexical information, but this distinction is not reflected in people's beliefs about what will be memorable.
The study is devoted to the analysis of the impact of globalization and digital technologies on the transformation of language norms in modern digital discourse. The relevance of the work is due to the growing role of digital communication, in which language becomes a key factor of cultural identity and social integration. The goal is to identify interdependencies between the level of digital maturity of states, the intensity of language hybridization and the dynamics of sociolinguistic variability. The object of the study is the digital language space of the global communication environment. The methodology is based on a combination of systemic, institutional, econometric and corpus-linguistic approaches using official statistical databases and corpora of digital texts for 2015–2024. The results show that during this period ICT Development Index increased from 5.32 to 7.41 points, DESI Index – from 48.7 to 70.4 points, and the share of Internet users – from 58.2% to 84.3%. At the same time, the number of languages with digital representation increased from 312 to 387, and the share of English-language content decreased from 55.1% to 49.3%. The hybridization index increased from 0.42 to 0.61, clearly indicating the establishment of a polycentric mode or multi-node digital discourse. Hybridization is both method and degree: different languages or their structural elements used in the same message (verbal or symbolic), as well as fully/partially mixed code messages; hashtags expressed in different linguistic forms/fonts, etc. (multimedia message). This means how various linguistic components are arranged within one structural unit up to multimedia messages comprising codes written with different fonts‐forms on various levels of hybridity. The econometric model registered a very high positive correlation between digital maturity and hybridization (r=0.82).
The integration of generative artificial intelligence into everyday interpersonal communication - through AI-assisted writing, real-time translation, automated summarization, and conversational suggestion tools - represents a structural transformation in the mediation of human-human communication. Unlike earlier communication technologies that mainly transmitted or stored human-generated content, generative AI actively co-produces communicative output through a system architecture that includes user interfaces, prompt and context management, large language model inference, personalization modules, disclosure controls, and post-generation human editing. This paper examines AI-mediated interpersonal communication (AIMIC) as both a social phenomenon and an implementable communication system. It analyzes three interconnected effects: linguistic convergence, redistribution of communicative labor, and challenges to relational authenticity. The revised AIMIC framework links these effects to concrete implementation layers in email assistants, messaging platforms, translation tools, collaborative writing systems, and organizational communication software. The paper argues that responsible AIMIC design requires preserving human communicative agency, making AI involvement contextually transparent, monitoring linguistic and relational impacts, and embedding governance mechanisms into the communication pipeline. The paper concludes with practical examples, evaluation recommendations, and design principles for platform developers, educators, organizations, and policymakers.
The authors created a hybrid quantum-classical model for efficient sentiment classification based on Frozen Distilled Bidirectional Encoder Representations from Transformers (frozen dBERT), supervised dimensionality reduction, variational quantum learning, and residual integration. The binary Stanford Sentiment Treebank had a total of 67,349 training observations. Following contradiction, duplication and overlap removal, 66,971 observations were left, of which 60,274 were used for training, and 6,697 for internal validation. There were 872 observations on the official development partition that were not touched. The embeddings, which had a dimension of 768, were mean-pooled to 4, 6, or 8 dimensions. The six-dimensional bottleneck was chosen and represented in the variational circuit with six qubits. The selected circuit was composed of 12 completed trials and 14 pruned trials, and had 3 layers, 36 quantum parameters, 52 logical gates and a circuit depth of 16. The classical six-dimensional bottleneck achieved 0.8360 accuracy, 0.8427 harmonic-mean classification score, 0.9184 receiver operating characteristic area under the curve and 0.6724 Matthews correlation coefficient on the official development partition. It was able to lower the dimensionality of the representation by 99.219 percent and obtain a model-head latency of 0.000607 milliseconds per observation. The quantum-only classifier had an accuracy of 0.6663 and a harmonic-mean classification score of 0.6914, whereas the residual hybrid kept a zero-quantum contribution. The results show efficient classical compression with no quantum advantage for the simulation of six-qubit states.
The advent of computational linguistics has fundamentally transformed the analysis of text complexity and the preservation of cultural heritage within the Indo-European language family. This article explores how computational methodologies, including natural language processing (NLP), machine learning, and phylogenetic analysis, enhance our understanding of phonetic, morphological, syntactic, and semantic complexities across diverse Indo-European languages. By leveraging digital corpora, treebanks, and automated annotation tools, researchers can uncover historical linguistic relationships, democratize access to underrepresented languages, and support revitalization efforts. Furthermore, the digitization and contextualization of ancient texts facilitate cross-cultural scholarship and safeguard endangered linguistic traditions.
This repository contains GSD-NP and GSD-DiNoS, both derived from Universal Dependencies' (UD) GSD Treebank. GSD-NP (.conllu) is a subset of UD-GSD and comprises its simplex noun phrases (NP): Common nouns (NN/NOUN) and their direct dependents (determiners, adnominal adjectives, nmods, adpositions, adverbs). It consists of 49,425 NPs (119.0k tokens) and has an improved feature annotation coverage (gender, case, number). Breaking with UD annotation, a total of 3,649 APPRART tokens were reconstructed in GSD-NP to restore the original orthographic forms. GSD-DiNoS (.json) is a custom data-driven lexion-like data structure built on GSD-NP, which aggregates NPs with the same head lemma. For each lemma, absolute frequencies of the lemma and its word forms are captured. Moreover, each occurrence feeds into three areas of interest within the word form entry: morphosyntactic features in isolation (gender, case, number), in combination with groups of dependents (collocations), and in combination with the syntactic function (dependency relations). GSD-DiNoS spans 17,433 unique lemmas and 20,190 unique word forms, stemming from 49,416 NPs. Lemmas were relemmatised to assign unique lemmas to nominal compounds, a highly productive and often lexicalised construction in German.
Government hiring processes face significant challenges in efficiently and equitably matching qualified candidates with appropriate positions while managing high application volumes. This paper presents an Intelligent Text Mining Framework for decision support in government hiring that integrates lexical and semantic similarity pathways through a weighted fusion mechanism. The framework processes resume and job description corpora using parallel TF-IDF (lexical) and Sentence-BERT (semantic) analyses, combines scores via a tunable parameter α, and ranks candidates per job description. A comprehensive evaluation on a dataset of 100 resumes and 10 job descriptions demonstrates that the hybrid approach achieves a mean combined similarity score of 0.642 ± 0.113 with high reliability (split half correlation r=0.891, p<0.001). The automated pipeline reduces screening time by 99.97% compared to manual review, reclaiming approximately 5.2 person months of effort per 1,000 comparisons. Using non-sensitive proxy variables like resume length and professional category, rigorous fairness tests show no disproportionate impact (80% rule ratio = 0.858) and no statistically significant bias between groups (Kruskal Wallis p=0.543). The system contains an AI dashboard that shows how scores are spread out, how the best candidates rank, and how big the skill gaps are. This helps hiring supervisors keep track of what's going on. The results suggest that the dual path method is a solid balance between precision and recall, helps choose candidates equitably, and is a scalable, auditable way to hire people in the public sector. This study provides a proven, open-source technology that improves government recruiting by making it more efficient, fair, and open, while still allowing for human monitoring and following ethical hiring norms.
The article examines the characteristics of stability and variability of the literary language norm based on Uzbek and English linguistic materials from a comparative sociolinguistic perspective. It analyzes the views of the Prague Linguistic School, Russian and Uzbek theories of language norms, and Western sociolinguistics concerning the antinomy of linguistic norms, substantiating the uneven distribution of stability across different levels of the language system. Internal and external factors contributing to normative changes in orthography, vocabulary, and grammar are identified, and the stages of norm shifts occurring through competition between linguistic variants are described. The study demonstrates that orthographic rigidity in English coexists with grammatical and lexical flexibility, whereas in Uzbek, the grammatical core retains its stability even under conditions of script instability.
This article provides a short overview of several key areas of digital classical scholarship where different types of open data from the Perseus Digital Library have been reused, enhanced and in some cases been reused by Perseus in turn. After providing some brief context regarding the history of and need for open data in the study of texts in classical languages it explores three areas of the reuse of Perseus data: TEI-XML texts, morphological and treebank data, and named entity data.
The author continues the research begun in previous articles («Philosophy and Culture», 2023, no. 11; «Philosophy and Culture», 2024, no. 7). The subject of this study is the Christian discourse of the British television series «Robin of Sherwood» (1984 – 1986). The author examines the problem of violence in the series' cinematic text, as well as in Russian-language literary and historical online resources created by its audience. Viewers' interpretations of violent episodes are compared from the perspective of various value-normative systems, including the norms of Soviet education, the laws of market capitalism, the principles of modern civil society, the mores of medieval European society, and the ideals of Christian ethics. The methodological basis of this work is a philosophical and hermeneutic approach, which utilizes a hermeneutic interpretation of the artistic text, incorporating elements of historical and cultural, contextual, intertextual, and lexical-syntactic analysis. An axiological approach is also employed to identify the value hierarchy within the worldviews of the film's creators and its viewers. The sociocultural concept of E. Durkheim is used as a paradigm explaining the nature of violence. A high degree of congruence is revealed between the values of Christian humanism and the ethical principles that underpinned the socialization of the film's Russian-speaking audience in the 1970s and 1980s. The most active discussions in the literary internet space revolve around episodes of violence by positive characters, primarily Robin Hood, against a helpless opponent. It is concluded that Robin Hood is depicted in the film as a sovereign possessing legitimate political power, one of whose prerogatives is the right to use violence. At the same time, the series' narrative demonstrates a «character arc» – the consistent development of the protagonist's self-awareness toward the norms of Christian morality.
The article examines phonetic variation in contemporary Ukrainian speech in relation to linguistic norms and communicative practices. The relevance of the study is determined by ongoing processes of language change, driven by sociocultural transformations, media environments, and intensified language contact. Special attention is paid to the variability of pronunciation, stress patterns, intonation models, and the adaptation of foreign elements in modern speech. The aim of the research is to identify the main types of phonetic variation and establish their correlation with the codified norms of the Ukrainian language. The methodological framework is based on the descriptive method, elements of phonetic analysis, and a functional approach to language phenomena. The empirical material encompasses examples from both spoken and written discourse, precisely media texts, public speech, and everyday communication. The results verify that phonetic variation is an inherent feature of contemporary Ukrainian speech and manifests itself at distinctive levels, from segmental to prosodic. The study reveals tendencies toward simplification of pronunciation, variability in stress placement, and fluctuation in intonation patterns. It is contended that communicative practice often precedes normative codification or contributes to its transformation. The study concludes that phonetic variation reflects the dynamic nature of language development and plays a pivotal role in shaping modern linguistic norms. Further research prospects comprise experimental phonetic analysis and the investigation of media influence on phonetic processes.
We present a Universal Dependencies (UD) annotated dataset of the Book of Ezra from the Old Georgian Oshki Bible, based on the TITUS edition. Our semi-automated workflow leverages transfer learning from Modern Georgian (UD_Georgian-GNC) and few-shot prompting of LLMs, followed by manual corrections by human linguists. We discuss the challenges of creating UD-conformed guidelines tailored for Old Georgian’s unique linguistic structure and the limitations and errors of LLM-based annotation. This work aims to establish a gold standard for Old Georgian UD, laying the groundwork for future morphosyntactic analysis of Old Georgian.
Abstract Objectives This study aimed to investigate the effects of brief mindfulness meditation on emotion regulation and its underlying brain functional network mechanisms in college students using functional near-infrared spectroscopy (fNIRS), thereby providing neuroscientific evidence for mindfulness-based interventions targeting emotional health. Method A total of 59 college students were randomly assigned to either a brief mindfulness meditation group (n = 29) or a control group (n = 30). Participants in the mindfulness group completed a 15-min mindfulness breathing intervention, whereas participants in the control group received safety education. Before and after the intervention, all participants performed an emotion regulation task involving negative viewing and cognitive reappraisal conditions. Changes in oxygenated hemoglobin (Oxy-Hb) concentration were recorded using a 40-channel fNIRS system across eight regions of interest, including the left dorsolateral prefrontal cortex (L-dlPFC), right dorsolateral prefrontal cortex (R-dlPFC), left frontopolar area (L-FPA), right frontopolar area (R-FPA), left premotor cortex (L-PMC), right premotor cortex (R-PMC), left primary somatosensory cortex (L-S1), and right primary somatosensory cortex (R-S1). Brain functional connectivity was constructed based on phase-locking value (PLV), and graph-theoretical metrics were subsequently calculated to characterize topological properties of the functional network. Results Behavioral results indicated that, compared with the control group, participants in the brief mindfulness meditation group exhibited significantly lower emotional valence ratings during the negative viewing condition (p < 0.001). Functional connectivity analyses revealed significantly increased connectivity between the L-dlPFC and L-FPA, L-dlPFC and R-PMC, and L-dlPFC and L-PMC following the mindfulness intervention ( p < 0.05). Furthermore, graph-theoretical analyses demonstrated significantly increased global efficiency, as well as enhanced local efficiency and clustering coefficient of the L-dlPFC in the mindfulness group ( p < 0.05). Conclusions Brief mindfulness meditation may effectively enhance functional integration within large-scale brain networks and improve neural information-processing efficiency in college students. These findings provide important neurophysiological evidence supporting the beneficial effects of mindfulness interventions on emotion regulation.
This article examines the role of Shakespearean translation and theatrical performance in fostering cultural exchange between English and Uzbek cultures. Shakespeare’s drama, rooted in the historical, linguistic, and philosophical context of early modern England, undergoes significant transformation when introduced into the Uzbek cultural space. Through careful translation and stage adaptation, Shakespeare’s works are reinterpreted to align with local linguistic norms, aesthetic traditions, and cultural values. This study argues that translation and theatre act as complementary mechanisms of cultural mediation, facilitating intercultural dialogue while preserving the universal themes of Shakespearean drama. The analysis demonstrates that these processes do not merely transmit literary content but actively construct shared cultural understanding.
This article examines the linguistic adaptation of aviation terminology in the Uzbek language. Special attention is given to phonetic, lexical, grammatical, structural, and semantic adaptation processes of aviation terms borrowed mainly from English through direct international communication and technological development. The study analyzes the mechanisms through which foreign aviation terms become integrated into Uzbek linguistic norms and professional discourse. Comparative, descriptive, structural, and semantic analysis methods are employed to identify the main adaptation patterns. The research demonstrates that aviation terminology in Uzbek undergoes systematic phonetic and morphological modifications while preserving technical precision and communicative effectiveness. The article also discusses the role of globalization, international aviation standards, and professional communication in the development of modern Uzbek aviation vocabulary.
This study investigates the grammaticalization of the Romance temporal construction ‘to have plus Perfect Participle’, employing a multifaceted approach that integrates theoretical linguistics, corpus analysis and computational methods. It critically assesses the potential of linguistic annotation pipelines and Neural Language Models to contribute to diachronic linguistic research, specifically in tracking the evolution of habeo into an auxiliary verb. The research explores whether automatic annotation models can accurately capture the early stages of grammaticalization, identifies the syntactic conditions that facilitate auxiliary annotation, and evaluates whether pre-trained models can reliably distinguish between the auxiliary and lexical uses of habeo. A corpus of Latin sentences, spanning the 3rd-5th and 8th-10th centuries, was compiled and processed using UD Pipe, revealing challenges in parsing related to data scarcity in the training treebank. The study also employs Latin BERT, a Neural Language Model, to examine the model's ability to differentiate between lexical and proto-grammatical uses of habeo. Preliminary results demonstrate that the Perfect Participle’s behavior became increasingly verbal over time, which is in line with the grammatical use of habeo. Furthermore, the study identifies a shift in the syntactic context of auxiliaries, with later periods showing a greater propensity for auxiliary use, which aligns with the observed model predictions. This work highlights the value of computational linguistic tools in diachronic studies, while also advancing our understanding of the syntactic features that characterize grammaticalization.
This registered secondary-data analysis tests whether clauses negated by 弗 have fewer overt syntactic objects than clauses negated by 不 when the same verb lemma occurs with both negators in the same transmitted work. It uses the frozen Kyoto Classical Chinese Universal Dependencies treebank v2.18, explicitly discloses exploratory counts already examined, and specifies an exhaustive extraction, matched-stratum test, and manual audit of apparent exceptions. The syntactic distribution can bear on accounts of 弗 as an incorporated-object form, but it cannot by itself establish the historical etymology of 弗. The analysis will not infer a time trend from the composite transmitted texts.
The Core Idea Classical combinatorics counts the number of ways to partition a set of n elements into non-empty blocks. That count is the Bell number B(n), and it depends only on how many elements there are, never on how they relate to each other. This manuscript introduces a strict refinement. Given a connected graph G on n labeled vertices, define a connected set-partition as a partition of V(G) in which every block induces a connected subgraph. The geometric partition function is p₍𝔾₎(G) = |{ {V₁, …, Vₖ}: each G[Vᵢ] is connected }| This count depends on G, not just n. Two graphs on the same number of vertices can have wildly different geometric partition counts. The entire theory follows from taking that dependence seriously. Main Theorems Theorem 1 (Subgraph Lattice Theorem). If G and H share the same vertex set and E(G) ⊆ E(H), then p₍𝔾₎(G) ≤ p₍𝔾₎(H). Adding edges can only increase the connected partition count, never decrease it. Theorem 2 (Tree Evaluation). For any tree T on n vertices, p₍𝔾₎(T) = 2ⁿ⁻¹. Every tree on n vertices, regardless of shape, has exactly the same geometric partition count. This is the universal floor for connected graphs. Theorem 3 (Complete-Graph Evaluation). For Kₙ, p₍𝔾₎(Kₙ) = B(n), the n-th Bell number. When every possible edge is present, every set-partition is automatically connected, so the geometric count recovers the classical count. This is the universal ceiling. Theorem 4 (Cycle Closed Form). For the cycle graph Cₙ with n ≥ 3, p₍𝔾₎(Cₙ) = 2ⁿ − n. This matches OEIS sequence A000325. Corollary (Sandwich Hierarchy). For every connected graph G on n vertices with n ≥ 4: 2ⁿ⁻¹ = p₍𝔾₎(Tₙ) < p₍𝔾₎(Cₙ) = 2ⁿ − n < p₍𝔾₎(Kₙ) = B(n) Both inequalities are strict. Trees sit at the floor, complete graphs at the ceiling, and the cycle sits strictly between them. What the Numbers Look Like For cycles versus trees versus complete graphs: n = 5: Tree = 16, Cycle = 27, Complete = 52 n = 6: Tree = 32, Cycle = 58, Complete = 203 n = 7: Tree = 64, Cycle = 121, Complete = 877 n = 8: Tree = 128, Cycle = 248, Complete = 4,140 n = 10: Tree = 512, Cycle = 1,014, Complete = 115,975 The gap between floor and ceiling grows super-exponentially. At n = 10, the complete graph has 227× more connected partitions than a tree on the same vertices. At fixed n, different topologies produce sharply different counts. At n = 6 for example: P₆ = 32 < C₆ = 58 < Ladder = 74 < Fan = 89 < K₂,₄ = 96 < Prism = 114 < W₆ = 118 < K₆ = 203 The ordering tracks graph density with Spearman ρ = +0.955 (p < 10⁻⁶) across all tested sizes n = 4 through n = 10. The Connection Curvature Define the partition connection on cycles as Γₙ = p₍𝔾₎(Cₙ₊₁) / p₍𝔾₎(Cₙ). The curvature κₙ = Γₙ − 2 measures deviation from the flat tree-growth rate: κₙ = (n − 1) / (2ⁿ − n) This is strictly positive and monotonically decaying to zero: n = 3: κ = 0.4000 n = 5: κ = 0.1481 n = 7: κ = 0.0496 n = 10: κ = 0.0089 The cycle approaches tree-like growth exponentially fast. The successive ratio κₙ / κₙ₋₁ converges toward ½ from above. Empirical Validation Program The deposit includes a preregistered validation suite with predeclared falsifiers, frozen metrics, and external ground truth. All outcomes, including failures, are retained as first-class results. Test 1 (OEIS Path Recursion). Brute-force enumeration confirms p₍𝔾₎(Pₙ) = 2ⁿ⁻¹ for n = 1 through 20. The sequence matches OEIS A000079 (powers of 2) exactly. Algebraic DP and direct bitmask enumeration agree at every tested value. Test 2b (Same-n Topology Comparison). At each n from 4 to 10, connected graphs on the same vertex count were compared. Spearman rank correlation between vertex connectivity and p₍𝔾₎ is ρ ≥ 0.83 at every tested n. Pooled edge-density correlation: ρ = 0.955 with p < 10⁻⁶. Negative control (max degree): ρ ≈ 0.35, not significant at any n. The lattice theorem's monotonicity prediction holds empirically, and counterexamples to vertex-connectivity ordering exist only between spanning-subgraph-incomparable pairs, exactly as the theory permits. Test 4 (Internet Autonomous Systems, 2010–2024). Using 180 monthly CAIDA BGP snapshots: AS count grew from 33,778 to 77,495 while edges grew from 93,998 to 501,473. The edge-succession to vertex-succession ratio Sₑ/Sᵥ increased with a linear slope of 3.68 per year, R² = 0.88, p = 2.23 × 10⁻⁷, Spearman ρ = 0.968. The pre-registered threshold was a slope above 0.05. The observed slope is 73× that threshold. Edge growth dominates vertex growth in a real infrastructure graph, consistent with the lattice theorem's prediction that denser graphs have exponentially more connected partitions. Test 5 (OpenAlex Citation Subgraphs). Among 10,000 co-citation subgraphs from 2020–2024: only 3 out of 10,000 were path graphs (0.03%). Among connected subgraphs, paths were 0.47%. For n ≥ 7, zero path graphs appeared among 527 connected subgraphs. Real citation networks avoid the tree floor almost entirely. Test 6/6b (Universal Dependencies Treebanks). The primary test fired its falsifier: English had 19.63% path-topology sentences, exceeding the 10% threshold. However, the pre-registered follow-up (Test 6b) conditioning on n ≥ 5 tokens found all six languages below 10%, with a maximum of 1.57% (Russian). At n ≥ 7 tokens, all languages drop below 0.15%. Short sentences are path-like; longer sentences are not. Test 7 (C. elegans Connectome). 279 neurons, 1,961 synapses. The biological network sits near the tree floor: 81% of 4-vertex connected induced subgraphs achieve p₍𝔾₎ = 2ⁿ⁻¹ exactly. KS tests showed no significant difference from Erdős–Rényi at matched density. Pre-registered hypothesis not supported. The sandwich bound is tight in sparse biological networks. Test 10 (Crystallographic Space Groups). Commutation graphs of all 230 space groups tested against a Ramsey proxy bound. The bound holds for 230/230 groups (100%). Saturation target (≥ 3 groups at n = 5 with zero triangles and zero independent triples) was not achieved. Verdict: partial. Tests 2, 8 (Falsifier Fired). Cycloalkane ring strain shows no monotonic observable matching p₍𝔾₎ (Pearson r = −0.11, p = 0.79). KEGG metabolic pathways achieved 14/18 hits for the Sₑ ≥ 2 × Sᵥ criterion versus a target of 15. Both failures are preserved in the record. Validation Ledger Supported: Tests 2b, 4, 5, 6b (4 of 10 completed tests) Falsifier fired: Tests 1 (by design), 2, 6, 8 (4 of 10) Not supported: Test 7 (1 of 10) Partial: Test 10 (1 of 10) Blocked/Deferred: Tests 9, 11 Negative and null results are first-class outcomes, not omissions. The program-level verdict is assessed across the full ledger, not cherry-picked from successes. Limitations Some tests depend on external APIs (OpenAlex, CAIDA, KEGG) whose upstream data evolves. Test 9 is blocked by endpoint observability constraints. This archive captures a dated snapshot; future reruns under changed data conditions should be treated as independent replications, not as invalidations. Citation Cite the Zenodo DOI for this archived package and the manuscript title. If reusing specific test scripts or results, cite the relevant script and JSON artifact in your methods section.
Sparse Mixture-of-Experts (MoE) routers commonly use the same scores both to select experts and to weight their already-computed outputs. We study whether these two roles, dispatch and aggregation, should be coupled. On pretrained OLMoE-1B-7B, we keep selected Top-8 expert IDs, expert computation, and total selected router mass fixed and change only within-set aggregation. A structured oracle improves full-horizon cross-entropy by 0.0160 +/- 0.0039 across three seeds; the router's top-scored expert is the counterfactual-best vertex only 17.2% of the time, with router-utility Spearman 0.030. We therefore train Fixed-Dispatch Adaptive Aggregation (FDAA), a 301K-parameter post-compute head optimized directly with the language-modeling objective while freezing the backbone, router, and experts. On OLMoE, FDAA improves fresh WikiText-103 test by Delta CE = -0.1523 +/- 0.0031 across three seeds, and mixed-domain training gives robust gains on WikiText-103, C4, and held-out Penn Treebank under frozen confirmatory evaluation. We also replicate the fixed-dispatch audit on DeepSeek-V2-Lite, which uses Top-6 routed experts plus shared experts. Best-vertex headroom remains significant on WikiText and C4, while router Top1 identifies the best selected expert in only 12.5% and 16.7% of audited examples. In a one-seed mixed-domain replication, FDAA improves locked WikiText and PTB, while C4 is statistically neutral. These results support a cross-architecture distinction between expert selection and expert commitment.
We show that emotion vectors in LLMs are organized by a two-dimensional valence-arousal (VA) subspace exhibiting circular geometry. Through principal component decomposition and ridge regression, we recover meaningful VA axes underlying emotion steering vectors whose projections correlate with human affect ratings across 44,728 words. Steering along these axes produces monotonic control over the affective properties of generated text, and further affords bidirectional control over multiple downstream behaviors (refusal and sycophancy) from a single subspace. These effects replicate across Llama-3.1-8B, Qwen3-8B, and Qwen3-14B. We propose lexical mediation to explain why these effects and prior emotionally framed controls work: refusal and compliance tokens occupy distinct VA regions, and VA steering directly modulates their emission probabilities.
Facial expression recognition (FER) in naturalistic settings is constrained by label ambiguity and variability in stimulus-response alignment. Adopting a data-centric perspective, this study examined whether emotional intelligence (EI)-stratified training data influence FER performance by treating EI as a qualitative factor associated with affective data consistency. Naturally elicited facial expressions were collected in a controlled emotion induction experiment with subjective arousal and valence ratings. Using response-driven labeling, neutral ratings were retained as indicators of ambiguity. Participants were grouped into High and Low EI based on the alignment between subjective evaluations and outputs from a pretrained affect estimator. Identical binary classifiers for arousal and valence recognition were trained while varying only the training data composition and evaluated across baseline, unambiguous, and ambiguous test sets using independent training repetitions with repetition-level statistical aggregation. EI-stratified training was associated with statistically detectable, context-dependent performance differences: group effects were observed primarily under baseline conditions and, to a lesser extent, under ambiguous conditions, whereas no reliable differences emerged under unambiguous conditions. Pooled discrimination differences were modest, but item-level analyses identified significant differences in classification correctness in specific task-condition combinations. Comparable patterns were observed across alternative backbone architectures. These findings indicate that FER performance in naturalistic contexts is influenced not only by model architecture but also by the statistical structure and internal coherence of the training data, supporting EI-informed data selection in ambiguity-prone scenarios.
Background Internalizing psychopathology is thought to be characterized by excessive responding to negative stimuli. Measuring this in the lab – e.g., via amygdala response and emotion ratings elicited by pictures – can provide insight into in-the-moment, time-limited emotional response in a controlled environment. Ecological momentary assessment (EMA) of everyday emotional experience can provide complementary measurement of emotion dynamics in response to real-world stimuli and events. When used together, these lab-based and everyday emotional assessments may account for additional variance in the development and worsening of psychopathology. Methods Seventy-five participants with mixed internalizing symptoms (50 female, 24 male, 1 other) were recruited from the community. fMRI BOLD signal and subjective valence ratings were acquired while participants viewed emotional pictures. Affective inertia for negative affect (NA) and positive affect (PA) was calculated using multilevel modeling of EMA data collected in participants’ daily lives. Internalizing symptoms were assessed at study entry and one year later. We tested how fMRI amygdala BOLD, picture ratings, and their interaction with affective inertia would predict internalizing symptoms one year later. Results In individuals with higher NA inertia only, greater negative > neutral fMRI BOLD and negative picture valence ratings predicted increases in internalizing symptom severity one year later. Hierarchical regression analyses further demonstrated that neural and subjective measures accounted for unique variance in symptom change. No significant results emerged for PA. Conclusions Individuals with heightened negative emotional reactivity for whom everyday negative emotion is also resistant to internal and external change may be at risk for increasing internalizing psychopathology.
Negative biases in emotional processing are central to cognitive models of Major Depressive Disorder (MDD), yet it remains unclear whether negatively biased valence attribution covaries with depressive symptoms or represents a more stable vulnerability. In this longitudinal study, we examined emotional valence ratings during a subliminal affective priming task in n = 232 MDD patients and n = 496 healthy controls (HC's) across multiple follow-up assessments, yielding 1395 observations. Multilevel models tested effects of depressive symptom severity, diagnostic group (MDD vs. HC), affective prime (happy, sad, neutral), and their interactions. To separate within-person fluctuations from between-person differences, depressive symptom severity was decomposed (Mundlak within-between specification). Exploratory analyses examined whether long-term symptom trajectory clusters were associated with emotional bias. Greater depressive symptom severity was associated with more negative valence ratings. Within-between analyses indicated that this association was primarily driven by within-person symptom fluctuations, indicating more negative ratings when symptom severity exceeded individual average. In contrast, diagnostic group, prime condition, and symptom severity × prime interactions did not significantly predict valence ratings, and no group differences emerged between HC's and remitted MDD patients. Exploratory trajectory analyses suggested that patients with recurrent symptom patterns may show more negative valence ratings, although cluster stability was limited and findings should be interpreted cautiously. Overall, the results suggest that negative valence attribution in MDD is more closely linked to current depressive symptom severity than to diagnostic status alone. Longitudinal approaches separating within- and between-person processes may help clarify when negative biases emerge and persist during depression progression.
Short-form video platforms have become central to multimedia information dissemination, where comments play a critical role in driving engagement, propagation, and algorithmic feedback. However, existing approaches -- including video summarization and live-streaming danmaku generation -- fail to produce authentic comments that conform to platform-specific cultural and linguistic norms. In this paper, we present LOLGORITHM, a novel modular multi-agent framework for stylized short-form video comment generation. LOLGORITHM supports six controllable comment styles and comprises three core modules: video content summarization, video classification, and comment generation with semantic retrieval and hot meme augmentation. We further construct a bilingual dataset of 3,267 videos and 16,335 comments spanning five high-engagement categories across YouTube and Douyin. Evaluation combining automatic scoring and large-scale human preference analysis demonstrates that LOLGORITHM consistently outperforms baseline methods, achieving human preference selection rates of 80.46\% on YouTube and 84.29\% on Douyin across 107 respondents. Ablation studies confirm that these gains are attributable to the framework architecture rather than the choice of backbone LLM, underscoring the robustness and generalizability of our approach.
Open-vocabulary image classification can recognize arbitrary textual categories but often fails to capture hierarchical relationships, for example, a Ragdoll cat should also be recognized as both cat and animal. Therefore, we introduce OpenHier, a comprehensive framework for open-vocabulary hierarchical classification that encompasses a hierarchical structure, a dataset, a benchmark, and a model. To capture hierarchical relationships, we construct a systematic real-world hierarchical structure based on the linguistic lexical database WordNet and Vision-Language Large Model. Building upon this structure, we develop a large-scale dataset with 4M annotated images, as well as a benchmark with 50K annotated images. Meanwhile, we design an evaluation pipeline to assess open-vocabulary hierarchical classification performance using several curated metrics. Furthermore, we propose a hierarchical consistency constraint method and multimodal alignment strategy to build a hierarchical classification model. Comprehensive experimental results demonstrate that our model achieves superior performance on both multi-label image classification and hierarchical image classification tasks in the open-vocabulary setting.
Pleonasms and pleonastic constructions represent complex processes of interaction between the logic of speech and its expressiveness, between the economy of linguistic means and stylistic overload. The problem of pleonasm or verbosity development represents important linguistic aspects, which are in the focus of present day linguistics. Its resolution touches upon fundamental issues of linguistics in general and semantics in particular, since in the modern context it is crucial to express oneself concisely, to form statements capable of objectively conveying information without elements that duplicate each other’s content, which is the essence of the language economy law. The relevance of the study is grounded in the high frequency of pleonastic constructions in modern communication and in the lack of systematic research addressing their functional potential. Pleonasm is often perceived as a hidden linguistic phenomenon, since in everyday communication it is rarely recognized as redundancy. Speakers intuitively accept pleonastic combinations as natural and even necessary, because their redundant elements are semantically integrated and no longer consciously perceived as repetitive. In this sense, pleonasm is explicit in structure but implicit in awareness: its redundancy is objectively present, yet subjectively unnoticed by language users. The paper considers the phenomenon of pleonasm through the prism of linguistic norms and linguistic usage, outlines its stylistic and syntactic potentials. The key object of the analysis of the problem is the conscious use of pleonastic compounds as means of expressiveness, as well as underlying semantic nature and pragmalinguistic force of constructions. Despite their prevalence in everyday speech, media discourse, literature, academic and public communication, pleonasms remain insufficiently described from a comparative and pragmalinguistic perspective. This paper responds to this problem by analyzing pleonasm as a dynamic phenomenon shaped by cognitive, historical, and communicative factors. Notwithstanding a previous scientific research on pleonasms as constructions that overload the text with unnecessary words, the study proved that pleonasticity performs important pragmatic functions in both the English and Ukrainian languages such as quantitative, emphatic and vocative functions in literary texts and poetic language where these stylistic techniques emotionally color the content.
This article examines the formation of borrowed, calqued, and hybrid terms in Uzbek chemical terminology. Chemical terminology is an essential component of the language of science and develops through the interaction of international scientific practice, the internal word-formation resources of the Uzbek language, and terminological units borrowed from other languages. The study focuses on the adaptation of chemical nomenclature to Uzbek linguistic norms, the assimilation of Russian-international terms, the creation of terms through calquing, and the emergence of hybrid terms based on international term elements. The analysis discusses borrowed units such as atom, element, proton, neutron, isotope, electrolysis, and catalyst; calqued naming patterns such as sodium sulfide, calcium sulfate, and potassium-sodium sulfate; and term elements such as hydro-, micro-, macro-, infra-, ultra-, iso-, and cryo-. The findings indicate that Uzbek chemical terminology is being refined through a balance between international scientific conventions and the grammatical, lexical, and word-formation norms of the national language.
Abstract This chapter provides an overview of the digital tools available to papyrologists, with a focus on resources for the linguistic annotation and analysis of papyrological data. The chapter also presents the aims, methods, and tools used by the PapyGreek project and introduces several key projects focusing on Ancient Greek that share common approaches with the authors’ own work. Special emphasis is placed on treebanking and its potential for facilitating scalability and comparative analysis across corpora. The chapter concludes with a set of scaled case studies demonstrating the potential of this approach for investigating language use in papyri, and outlines desiderata for the future development of digital resources related to papyri.
The increasing use of artificial intelligence (AI) in hiring raises significant ethical concerns, particularly regarding language-based biases. Although AI systems are promoted as tools to reduce human prejudice, research shows they can reproduce and even intensify existing inequalities. This qualitative study, based on secondary data, explores how resume language influences AI hiring decisions, focusing on linguistic patterns and the ethical implications of automated screening. Drawing on case studies and prior literature, it examines how algorithms interpret tone, action verbs, and culturally coded expressions. Findings indicate that AI tools often favor resumes aligned with dominant linguistic norms, such as industry jargon and assertive or masculine-coded language, putting underrepresented groups and non-native speakers at a disadvantage. The study also highlights how keyword filtering and biased training data reinforce systemic inequalities. Overall, it emphasizes the need for transparent, inclusive, and linguistically aware AI systems, calling for interdisciplinary efforts to ensure fair and equitable hiring practices.
Abstract Recent advances in text-to-music (TTM) generation have enabled controllable and expressive music creation using natural language prompts, yet the extent to which these systems faithfully convey intended emotions remains largely underexplored. In this study, we introduce AImoclips, a benchmark designed to evaluate emotion conveyance to human listeners in TTM systems using a dimensional valence-arousal framework. We constructed the dataset with 991 instrumental music clips from six TTM systems prompted with 12 emotion words spanning four valence-arousal quadrants. A total of 111 participants provided 6,162 valence and arousal ratings using a 9-point scale. The results revealed that all systems perform above chance in conveying quadrant-level emotion intent, yet overall accuracy remains limited. Notably, substantial model-dependent biases were present. Commercial systems tended to exhibit positive valence deviations, whereas open-source models more often produced outputs with a negative valence shift, with diverging arousal shifts. Furthermore, overall audio quality (Fréchet Audio Distance (FAD)) correlates with both valence and arousal, while text-audio alignment (Contrastive Language-Audio Pretraining (CLAP) score) primarily relates to valence. These findings highlight underlying challenges in conveying linguistic emotional semantics precisely through musically conveyed affect for future research on controllable and perceptually consistent music generation.
This study analyzes the impact of feminist policy in contemporary Germany on the stability of linguistic norms within political discourse. The article examines the process of integrating gender markers into the language, which is interpreted as a breach of traditional grammatical and orthographic norms in German. The research aims to identify linguistic markers that indicate artificial language transformation driven by new political agendas. The materials for this study include two policy documents published in 2023 by the Federal Foreign Office and the Federal Ministry for Economic Cooperation and Development of Germany, collectively comprising 235,000 characters of empirical data. Through qualitative and quantitative analysis, the study demonstrates how the conceptual core of feminist policy is expressed through non-canonical graphic and morphological means, leading to violations of the language norms in modern German. The authors conclude that the observed trend represents a form of politically motivated linguistic intervention, creating an alternative usage that contradicts codified norms.