Semantic Variation Research Articles

Affective Norms for German as a Second Language (ANGL2).

The present study introduces affective norms for a set of 880 German words rated by learners of German as a second language (L2), i.e., the Affective Norms for German as a Second Language (ANGL2). The database provides ratings across affective and subjective psycholinguistic dimensions. Besides valence and arousal ratings, ANGL2 features data on emotional prototypicality, which helps to identify emotion-label words and emotion-laden words. Moreover, the database includes two additional semantic variables: concreteness and familiarity. We observed similarities with previous studies, and the ratings provided by L2 speakers demonstrate characteristics that should be noted in studies involving bilinguals, including more moderate valence ratings, and a stronger correlation between valence and arousal, specifically for positive words. ANGL2 is the first set of affective norms that has been rated by L2 speakers for a language other than English. The set of norms is aimed to function as a resource for psycholinguistic experimental studies on the intersection between emotion and language among L2 speakers.

Dual-Attention Multiple Instance Learning Framework for Pathology Whole-Slide Image Classification

Conventional methods for tumor diagnosis suffer from two inherent limitations: they are time-consuming and subjective. Computer-aided diagnosis (CAD) is an important approach for addressing these limitations. Pathology whole-slide images (WSIs) are high-resolution tissue images that have made significant contributions to cancer diagnosis and prognosis assessment. Due to the complexity of WSIs and the availability of only slide-level labels, multiple instance learning (MIL) has become the primary framework for WSI classification. However, most MIL methods fail to capture the interdependence among image patches within a WSI, which is crucial for accurate classification prediction. Moreover, due to the weak supervision of slide-level labels, overfitting may occur during the training process. To address these issues, this paper proposes a dual-attention-based multiple instance learning framework (DAMIL). DAMIL leverages the spatial relationships and channel information between WSI patches for classification prediction, without detailed pixel-level tumor annotations. The output of the model preserves the semantic variations in the latent space, enhances semantic disturbance invariance, and provides reliable class identification for the final slide-level representation. We validate the effectiveness of DAMIL on the most commonly used public dataset, Camelyon16. The results demonstrate that DAMIL outperforms the state-of-the-art methods in terms of classification accuracy (ACC), area under the curve (AUC), and F1-Score. Our model also allows for the examination of its interpretability by visualizing the dual-attention weights. To the best of our knowledge, this is the first attempt to use a dual-attention mechanism, considering both spatial and channel information, for whole-slide image classification.

Negative concord by phase: multiple downward agree and the parametrization of edge features

Abstract We divorce Negative Concord (NC) among two or more n-items, which is invariably present in Romance languages, from the mutual exclusion or cooccurrence between clausal negation markers (CNMs) and other n-items. The standard minimalist treatment in terms of Agree with respect to a formal feature [n] (for [negation]) is maintained here. However, an EPP-like principle, conceived as a (parametrized) condition regulating the representation of [n] at phase edges is introduced to deal with CNMs. By divorcing the n-EPP condition from n-Agree, we eliminate Zeijlstra’s Upward Agree in favor of Multiple Downward Agree between a phase head probe endowed with Formal Feature FF and one or more elements endowed with the same interpretable feature FF in the WorkSpace. We argue that these theoretical moves are beneficial from an empirical point of view. Focusing on varieties spoken in Northern Italy, we argue that an exhaustive typology of Romance negative systems can be derived by the parametrized [n]-EPP. We claim that the present approach allows this variation to be accounted for within the syntax – specifically, no recourse is made to semantic variation.

Semantic variation and semantic change in the color lexicon

Abstract Color terms show a remarkable variation in their possible lexicalization patterns across different languages. In the literature, the interest has been especially to describe the color lexicon of a certain language and to determine whether it may abide or not by Berlin and Kay’s universal evolutionary sequence, e.g., whether a certain color denomination may be considered as a basic color term, or whether a certain color category is lexicalized by more than one basic color term, by which criteria, etc. It has not been established, however, which are the most common lexical sources of basic color terms on a comparative basis, and which semantic changes are more common in this semantic field. On the basis of data drawn from 70 ancient and modern Indo-European languages, we aim at answering precisely these research questions concerning the origin and the development of basic color terms. We discuss the various lexical sources of the basic color terms for white, black, red, green, yellow, blue, brown, gray, orange, pink, and purple, and we show the most important semantic changes leading to these color meanings. We also discuss to which extent these terms are likely to be inherited or borrowed. All this aims at being a contribution to the study of diachronic semantics.

Multivariate prototype representation for domain-generalized incremental learning

Deep learning models often suffer from catastrophic forgetting when fine-tuned with samples of new classes. This issue becomes even more challenging when there is a domain shift between training and testing data. In this paper, we address the critical yet less explored Domain-Generalized Class-Incremental Learning (DGCIL) task. We propose a DGCIL approach designed to memorize old classes, adapt to new classes, and reliably classify objects from unseen domains. Specifically, our loss formulation maintains classification boundaries while suppressing domain-specific information for each class. Without storing old exemplars, we employ knowledge distillation and estimate the drift of old class prototypes as incremental training progresses. Our prototype representations are based on multivariate Normal distributions, with means and covariances continually adapted to reflect evolving model features, providing effective representations for old classes. We then sample pseudo-features for these old classes from the adapted Normal distributions using Cholesky decomposition. Unlike previous pseudo-feature sampling strategies that rely solely on average mean prototypes, our method captures richer semantic variations. Experiments on several benchmarks demonstrate the superior performance of our method compared to the state of the art.

Addressing Semantic Variability in Clinical Outcome Reporting Using Large Language Models

Background/Objectives: Clinical trials frequently employ diverse terminologies and definitions to describe similar outcomes, leading to ambiguity and inconsistency in data interpretation. Addressing the variability in clinical outcome reports and integrating semantically similar outcomes is important in healthcare and clinical research. Variability in outcome reporting not only hinders the comparability of clinical trial results but also poses significant challenges in evidence synthesis, meta-analysis, and evidence-based decision-making. Methods: This study investigates variability reduction in outcome measures reporting using rule-based and large language-based models. It aims to mitigate the challenges associated with variability in outcome reporting by comparing these two models. The first approach, which is rule-based, will leverage well-known ontologies, and the second approach exploits sentence-bidirectional encoder representations from transformers (SBERT) to identify semantically similar outcomes along with Generative Pre-training Transformer (GPT) to refine the results. Results: The results show that the relatively low percentages of outcomes are linked to established rule-based ontologies. Analysis of outcomes by word count highlighted the absence of ontological linkage for three-word outcomes, which indicates potential gaps in semantic representation. Conclusions: Employing large language models (LLMs), this study demonstrates its ability to identify similar outcomes, even with more than three words, suggesting a crucial role in outcome harmonization efforts, potentially reducing redundancy and enhancing data interoperability.

Semantic Valence and Cultural Representation of Phraseological Units (Kazakh and English)

The rules for the compatibility of idiomatic units in the linguistic consciousness of native speakers are a particular problem that is related to the formation and evolution of the language norm. It is still unknown what is the catalyst for the transformation of language norms — usage or the laws and rules of the language. As a result of the cultural interpretation of a phraseological unit in the process of its use, the most important component of the cultural and linguistic meaning of the phraseological unit is formed, the content of which is the value-emotional attitude to what is happening. The focus of the article is the notion that phraseological valence - beyond grammatical structure to cultural and semantic dimensions. This perspective likely emphasizes the importance of cultural and semantic factors in shaping the meaning and usage of phraseological units, rather than purely grammatical considerations. Comparing English and Kazakh phraseological units with similar concepts but different cultural representations provides an insightful examination of how language reflects unique cultural perspectives and values. The article determines that there is a certain pattern in the semantic variation of phraseological units, and also compares the semantically equivalent and non-equivalent structure of English and Kazakh phraseological units. The purpose of the article is to prove that in the valence of phraseological units the role of the transitive meaning, rather than the literal meaning, predominates, and the grammatical construction has valency.

From words to senses: A sense-based approach to quantitative polysemy detection across disciplines

Interdisciplinary polysemy refers to the phenomenon in which a word is understood and used differently across disciplines or fields. Nuanced variations in word meanings pose a significant challenge to effective communication and knowledge integration, especially in academic and professional settings. To address these issues, we propose the Interdisciplinary Polysemy Detection (IPD) method, which streamlines the identification of polysemous words by making the process more automatic, objective and efficient. IPD not only identifies polysemous words but also quantifies the extent of semantic variation within and across disciplines, providing deeper insights into the scale and nature of meaning shifts in different contexts. The evaluation results demonstrate that IPD can identify interdisciplinary polysemy effectively, showing heightened sensitivity to subtle differences in word senses. This sensitivity makes IPD a useful tool for both researchers and educators, helping to improve communication across fields and contributing to the development of more tailored pedagogical resources. The broader implications of this method may extend to enhancing the understanding of semantic variation in interdisciplinary research and advancing knowledge transfer between specialised domains.

Syntactic and semantic variations in negative interrogatives: Contracted vs. uncontracted forms

This study investigates the forms and uses of negative interrogatives with auxiliary verb contractions in English, focusing on syntactic and semantic differences. Data was drawn from conversations in an English class at a university in Lhokseumawe, Indonesia, between a lecturer and students and students and students. The aim was to examine three types of negative interrogatives: Yes-No questions, negative rhetorical questions, and both marked (n’t) and unmarked (not) forms. The research employs a descriptive method with a synchronic time frame, utilizing distributional analysis, including immediate constituent analysis and mark reading techniques, to categorize and analyze the data. The findings reveal that negative interrogatives involving auxiliary verbs exhibit positive and negative presuppositions, with adjectival negative interrogatives showing variations between contracted and uncontracted forms. Adverbial negative interrogatives were found across all three types, indicating their versatility in forming questions. The study highlights that the choice between ‘not’ and ‘n’t’ affects the tone and emphasis of negative interrogatives, influencing the interpretation of responses. Contracted forms often carry a rhetorical or presumptive tone, while uncontracted forms are more formal and explicit. These variations show the importance of auxiliary verbs and negative markers in shaping the structure and meaning of negative interrogatives. This research contributes to the understanding of how negative interrogatives function in English, revealing the nuanced ways in which they impact communication and interpretation.

A Study on Internet Interactive Language Variations from the Perspective of Schema Theory

Language variation is a common phenomenon of language evolution in online interactive communication. Based on schema theory, this paper explores the forms, reasons, features and ritual functions of language variations in the context of Internet interaction. This study finds that the forms of Internet language variations are mainly embodied in: phonetic variation, semantic variation and grammatical variation, characterized by four rituals: cooperation, pluralism, universality and alignment. Internet language variations contribute to the construction of identity, relationship and order in communication. Through a systematic study of Internet language variations, this paper aims to provide some enlightenment for the ritual turn of pragmatics and the study of network linguistics from the perspective of cognitive linguistics.

The count-mass distinction in Hong Kong Sign Language: an intra- and cross-modal comparison

Abstract This article is in part an intra- and cross-modal comparison of the count-mass distinction (CMD) in Hong Kong Sign Language (HKSL). For the intra-modal analysis, HKSL data are compared to data found in Koulidobrova’s (2021) work on the CMD in American Sign Language/ASL (Koulidobrova, Elena. 2021. Counting (on) bare nouns: Revelations from American Sign Language. In Tibor Kiss, Francis Jeffry Pelletier & Halima Husić (eds.), Things and Stuff: The Semantics of the Count-Mass Distinction, 213–231. Cambridge: Cambridge University Press). According to Koulidobrova, the existing diagnostics used in spoken language research are insufficient to bring to light CMDs in sign languages. ASL’s CMD is visible in the failure to conjoin count and mass nouns and in the ungrammaticality of partitive constructions containing mass nouns. HKSL and ASL differ from each other in a few respects, among which the two aforementioned ASL diagnostic criteria. Regarding the cross-modal comparison, HKSL and ASL belong to different categories in Chierchia’s (2010) count-mass typology (Chierchia, Gennaro. 2010. Mass nouns, vagueness and semantic variation. Synthese 174. 99–149). Chierchia’s typology is based on CMDs in spoken languages. ASL is a number-neutral language (Type III), whereas HKSL is a number marking language (Type I) and thus patterns with languages like English and Dutch. The CMD in HKSL is visible in its failure to combine mass nouns directly with numerals and count adjectives without the intervention of a classifier. Based on my data analysis, I can furthermore argue that HKSL is a number marking language but that its plural number marking is realised through zero marking.

Transformer-Based Sequence Modeling Short Answer Assessment Framework

Automated subjective assessment presents a significant challenge due to the complex nature of human language and reasoning characterized by semantic variability, subjectivity, language ambiguity, and judgment levels. Unlike objective exams, subjective assessments involve diverse answers, posing difficulties in automated scoring. The paper proposes a novel approach that integrates advanced natural language processing (NLP) techniques with principled grading methods to address this challenge. Combining Transformer-based Sequence Language Modeling with sophisticated grading mechanisms aims to develop more accurate and efficient automatic grading systems for subjective assessments in education. The proposed approach consists of three main phases: Content Summarization: Relevant sentences are extracted using self-attention mechanisms, enabling the system to effectively summarize the content of the responses. Key Term Identification and Comparison: Key terms are identified within the responses and treated as overt tags. These tags are then compared to reference keys using cross-attention mechanisms, allowing for a nuanced evaluation of the response content. Grading Process: Responses are graded using a weighted multi-criteria decision method, which assesses various quality aspects and assigns partial scores accordingly. Experimental results on the SQUAD dataset demonstrate the approach’s effectiveness, achieving an impressive F-score of 86%. Furthermore, significant improvements in metrics like ROUGE, BLEU, and METEOR scores were observed, validating the efficacy of the proposed approach in automating subjective assessment tasks. Doi: 10.28991/HIJ-2024-05-03-06 Full Text: PDF

Neighborhood in Chinese lexicon: A megastudy analysis of lexical decision and naming of two-character Chinese words.

The present study examines the impact of neighborhood size (number of other two-character words sharing the same character at the same position) on Chinese lexical processing, along with its joint effects with variables such as character frequency, word frequency, and semantic transparency. Previous factorial experiments have yielded conflicting results that are difficult to reconcile with existing models (Li et al., 2015, 2017). To provide high-powered tests for these theoretically important effects on visual word recognition, we leveraged the megastudy approach and used linear mixed-effect analyses to investigate lexical decision and naming responses to a large pool of two-character Chinese words (N > 17,000) sourced from Tse et al.'s (2017, 2023) database. In all analyses we controlled for extraneous orthographic (e.g., stroke count), phonological (e.g., consistency), and semantic (e.g., transparency) variables. In addition to evaluating Li et al.'s (2015, 2017) models, we also investigated whether the parallel dual-route mechanism, which entails lexical access via whole-word or character decomposition-then-composition, could account for neighborhood size effect and its interactions in lexical decision and naming. Finally, we discuss the implications of our findings on the specificity of lexical effects with regard to character position and lexical processing task. (PsycInfo Database Record (c) 2024 APA, all rights reserved).

Network Analysis of Semantic Variations in Synonymous Verbs Across British and American English

Network Analysis of Semantic Variations in Synonymous Verbs Across British and American English

A Study on the Origins and Semantic Variations of ‘Modeng’

A Study on the Origins and Semantic Variations of ‘Modeng’

Analyzing the latent space of GAN through local dimension estimation for disentanglement evaluation

The impressive success of style-based GANs (StyleGANs) in high-fidelity image synthesis has motivated research to understand the semantic properties of their latent spaces. In this paper, we approach this problem through a geometric analysis of latent spaces as a manifold. In particular, we propose a local dimension estimation algorithm for arbitrary intermediate layers in a pre-trained GAN model. The estimated local dimension is interpreted as the number of possible semantic variations from this latent variable. Moreover, this intrinsic dimension estimation enables unsupervised evaluation of disentanglement for a latent space. Our proposed metric, called Distortion, measures an inconsistency of intrinsic tangent space on the learned latent space. Distortion is purely geometric and does not require any additional attribute information. Nevertheless, Distortion shows a high correlation with the global-basis-compatibility and supervised disentanglement score. Our work is the first step towards selecting the most disentangled latent space among various latent spaces in a GAN without attribute labels.

The effects of endogenous semantic variables during productive lexical retrieval: A Behavioral-Neural Dual Swinging Model (DSM)

Conceptual preparation is the very initial step in language production. Endogenous semantic variables, reflecting the inherent semantic properties of concepts, could influence the productive lexical retrieval by modulating both conceptual activation and lexical selection. Yet, empirical understandings on this process and underlying mechanisms remain limited. Here, inspired by previous theoretical models and preliminary findings, we proposed a Behavioral-Neural Dual Swinging Model (DSM), revealing the swinging process between conceptual facilitation and lexical interference and extending to neural resource allocation during these processes. To further test the model, we examined the joint effect of semantic richness and semantic density on productive word retrieval both behaviorally and neurally, using a picture naming paradigm. Results nicely support the DSM by showing that the productive retrieval is driven by the swinging between semantic richness-induced conceptual facilitation primarily managed in semantic-related regions and semantic density-induced lexical interference managed in control-related regions. Moreover, the conceptual facilitation accumulated from semantic richness plays a decisive role, mitigating interference from competitors as well as the neural demands in control-related regions.

Refining Environmental Sustainability Governance Reports through Fuzzy Systems Evaluation and Scoring

Environmental, Social, and Governance (ESG) reports have become essential tools for enterprises to showcase their commitment to sustainable development and social responsibility. However, discrepancies persist regarding the criteria, assessments, and ratings disclosed in these reports. Moreover, there is a need for more objective methods to determine the weight distribution of indicator items. This study introduces a novel approach utilizing semantic variables in fuzzy theory and a multiple logic fuzzy inference system to develop an ESG environmental management performance assessment model. Therefore, this paper aims to develop a novel approach utilizing semantic variables and a multiple logic fuzzy inference system to quantitatively evaluate the sustainable performance of an environmental management plan. This research also aims to ensure fair and objective assessment outcomes, providing valuable guidance for enterprises in implementing performance management strategies. Key aspects investigated include the impact of membership functions, the extended utilization of semantic variables and logical rules, a comparative analysis of traditional weight assessments, and the limitations of applying fuzzy theory. Through comprehensive discussions and calculations, it is evident that fuzzy theory offers considerable flexibility in application. By tailoring fuzzy rules and selecting appropriate membership functions, diverse application scenarios can be accommodated. The Fuzzy systems evaluation and scoring EMP model generates EMP evaluation scores ranging from 1.76 to 8.29 for Gaussian membership, 1.80 to 8.19 for Triangular membership-A, 1.92 to 8.00 for Triangular membership-B, and 1.81 to 8.19 for Quadrilateral trapezoidal membership, based on simulated rating scenarios using the semantic variables of completeness and feasibility. This approach successfully incorporates distribution logic from subjective membership degrees to evaluate EMP scores. The findings demonstrate that fuzzy theory enables the consideration of multiple factors and facilitates the provision of objective-level membership, underscoring its potential in addressing complex evaluation challenges. This study illuminates the versatility of the fuzzy system theory, with its applications poised to extend across various domains.

(Anti)Causativization of psych verbs in Spanish and Japanese

There is a notable typological contrast between psych verbs in Japanese and Spanish. Japanese derives Experiencer-Object verbs (e.g. yorokob-ase-ru ‘to please’) from specific Experiencer-Subject verbs (e.g. yorokobu ‘to become pleased’) via a morphological causativization. Spanish, on the other hand, presents so-called reflexive psych verbs (e.g. alegrarse ‘to feel happy’), most of which can be analyzed as outputs of an anticausativization from certain Experiencer-Accusative verbs (e.g. alegrar ‘to make happy’). Simply put, these languages derive psych verbs with procedures that reversely mirror each other. This paper will elucidate the characteristics of the causativization used to produce Japanese Experiencer-Object causatives and the anticausativization associated with Spanish Reflexive Psych Verbs and demonstrate that the typological contrast between Japanese and Spanish psych verbs results in semantic variation, e.g. differences in the entailment relation, absence/presence of ambiguity in negation, aspectual diversity. Semantic differences between psych verbs in these languages are ascribed to specific features of the (anti)causative operations employed to generate the predicates.

“Civilization State” Concept in the Works of Non-Western Scholars

The article discusses the main ideas and theories of non-Western countries researchers that reveal the concept of “civilization state”. As a methodological basis for the study of ideas developed within the framework of a non-Western tradition, the author uses the provisions of decolonial theory, allowing to overcome the clear and implicit dominance of attitudes and epistemological foundations inherent in European-American social science. The article draws attention to the critical perception of decolonial optics, which insists on the need to revise the epistemological foundations of Western thinking and the world order as a whole. The author analyzes the features of distribution and development of “civilization state” concept in a number of non-Western countries in the first quarter of the 21st century; the article presents a brief overview of emergence and evolution of the “civilization” concept, the diversity of semantic and conceptual variations of this concept is shown, the features of this process in the conditions of decolonization of knowledge were identified, an analysis of modern ideas and justification of the “civilization state” concept in China, India, Turkey, Egypt and in a number of African countries was carried out. The author draws conclusions that the concept of “civilization state”, being the Western origin concept, nevertheless, is perceived and develops within the framework of overcoming the West-centric theories of social development. In the conclusion the main advantages, threats and risks are presented that are associated with the adoption of the “civilization state” concept. Among the advantages are such as cultural and value consolidation of society; political mobilization; the basis for sovereignty and an independent policy; preservation and development of civilizational identity; the uniqueness of the civilization country as a source of further development. Challenges and risks include the following: the ambiguity of the new world order; contradictions between the concept of civilization and the concept of the national state; civilizational populism; ideas of isolationism; the risks of becoming the failed civilizational state. The article also recognizes the need for consolidated research work for the further theoretical development of the “civilization state” concept.