The paper presents a toolkit and an algorithm for assessing the lexical complexity of educational texts used in teaching Russian as a Foreign Language (RFL). In the context of educational digitalization, the relevance of this issue is driven by the need for objective, scalable tools to calibrate the complexity of learning materials. The study employs computational linguistics methods, including a custom Python script (process_word_lists) and large language models such as GLM 4.6, Grok 4 fast, Claude Sonnet, GPT-5, and Gemini 2.5 Pro. The research material comprises two text datasets: (1) a training set, which includes standardized RFL lexical minimums and 268 educational texts spanning levels A1–C1 (according to the CEFR), and (2) a test set consisting of 26 reading texts at levels A2–B1. Expert evaluation and statistical metrics—specifically Cohen’s kappa, Mean Absolute Error (MAE), ordinal accuracy, and nominal accuracy—were used to evaluate classification quality. The proposed algorithm enables highly reliable ranking of texts by CEFR difficulty levels. The revealed variations in the ability of large language models to assess RFL text complexity indicate high accuracy demonstrated by the GLM 4.6 and Grok 4 fast models. The developed algorithm and toolkit allow for the automated calibration of educational materials according to CEFR levels, enhance the reliability of expert evaluation, and expand the potential for developing adaptive educational resources. This functionality is in demand by both textbook authors and RFL test developers when selecting primary and calibrating secondary texts for textbooks and digital educational platforms. Future research prospects involve expanding the test dataset, analyzing texts at B2–C1 levels, and integrating new large language models.
This article examines the semantic and discursive features of generated texts created by large language models such as DeepSeek using German culture as an example. The topics of these generations primarily involve personal associations conditioned by the human perception system, specifically, sound associations with the concept “homeland.” This choice stems from the fact that such texts are closer to personality-oriented types of discourse and allow for a better understanding of the generation characteristics (in terms of the reliability of the information obtained from them, as well as the structural and semantic features of the texts), which explains the relevance of the study. The aim of the article is to identify linguistically specific means of representing the sonic image of the homeland in generated discourse using German culture as an example. The study material includes texts generated by the large language model DeepSeek. A survey of 78 native German speakers living in Germany was also conducted and asked the questions. The study uses a comprehensive methodology that includes elements of linguistic, conceptual, discourse, and contextual analysis. A comparative method is used to compare the generated sound images with the respondents’ answers. The author of the article believes that two components can be identified in the generated texts by DeepSeek: communicative and informative. The topic of the request and the mode of address (polite/informal) influence the style of presentation in both aspects. The development of the creative function of online communication (forums, comments) at the phonetic, lexical, and grammatical levels is reflected in the generated texts (especially in the communicative component), depending on the recipient’s request style and the linguistic resources incorporated into it. In answering questions, DeepSeek uses syntactic and semantic structures close to institutional discursive practices, namely, linguistic research, categorizing associative sounds into thematic groups using a classificatory approach. The style of presentation is poetic. For Zoomers, DeepSeek cites the sounds of gadgets as associative reactions characterizing the Homeland. No such associations were found in the Zoomer survey.
The digital era, marked by the mass introduction of neural machine translation systems, has given rise to a fundamental contradiction: high formal translation accuracy is increasingly accompanied by a deficit of recipient trust. In sensitive communicative domains (medicine, law, user interfaces), this contradiction becomes critical. The traditional “natural attitude” of the translator’s consciousness, which treats translation as an external object, leads to templating, algorithmic bias, and a loss of living contact with the recipient. A transition to a phenomenological attitude, in which translation is conceived as a correlate of the translator’s and the recipient’s consciousness, is required. The aim is to develop and empirically verify a four‑stage digital algorithm for the transition from the intentionality of distrust to the intentionality of trust, based on the phenomenology of E. Husserl and the concept of the “digital consumer”. The corpus comprised 320 digital texts from four discourse domains translated by 30 professional translators divided into a control (natural attitude) and an experimental (phenomenological algorithm) group. The experimental group underwent training including epoché and analogising apperception based on the “digital consumer” state. Comprehension accuracy, subjective trust (10‑point scale), translation time, and translators’ cognitive load (NASA‑TLX) were measured. Qualitative interviews with recipients were also conducted. Translations produced according to the phenomenological algorithm showed a statistically significant (p < 0.01) increase in comprehension accuracy from 74 % to 93 % and in subjective trust from 5.9 to 8.4 points. Translation time decreased by 24 %, translators’ cognitive load dropped from 55 to 34 NASA‑TLX points. The qualitative analysis revealed a substantial reduction in complaints about translations being “soulless” and an increased sense of the texts’ “humaneness”. The proposed digital algorithm, operationalising the phenomenological attitude, effectively bridges the gap between the formal accuracy of machine translation and the recipient’s need for trust, while simultaneously easing the translator’s work. The algorithm can be integrated into hybrid “human–AI” systems and into translator training programmes.
This article describes lexical, punctuation, orthographic, graphic features of professional communication in Russian-language aviation chats. The aim of the study is to identify and systematize these linguistic characteristics and to determine their functional role in the context of informal digital communication among aviation specialists. The research material comprises 500 messages and 334 specialized lexical units collected through a targeted selection process from two Russian-language aviation Telegram chats – “Aviaboltalka”, which focuses on technical issues and aircraft operations, and “DJI Lab”, which brings together pilots and operators of unmanned aerial vehicles. The following research methods were used in the study: the method of classification analysis, the method of stylistic analysis, the method of contextual analysis necessary to demonstrate the real functioning of the analyzed lexical units and sentences in aviation chats. The study found that lexical features manifest themselves in the active use professionalisms, and jargonisms. Syntactic features include the use of simple sentences and detachment. Punctuation features include the use of exclamation marks, ellipses, and violation of punctuation norms. Graphic features involve the use of emoticons as visual means of conveying emotions. The results of the study allow us to conclude that all identified linguistic features of professional communication in aviation chats are functionally conditioned by the operation of two leading pragmatic principles – minimizing the time expenditure and maximizing emotional impact. The identified linguistic features indicate a deep adaptation of professional speech to the norms of digital communication.
The article addresses the problem of developing professional post-editing competences in the context of neural machine translation dominance. The aim of the study is to provide theoretical substantiation and empirical verification of a two-level translation assessment model as a didactic tool. The research methods include the author’s experiment on translating culturally marked items via Google Translate and DeepL, two-level error analysis (technological diagnostics and communicative-functional assessment), quantitative and qualitative analysis of the obtained data, as well as comparative and contextual analysis of translation variants. The empirical part draws on authentic texts of Kuban regional character: the menu of the Cossack cuisine restaurant “Stan”, an announcement of a festival in the “Ataman” ethno-complex, and a fragment of Kuban balachka. All items were translated via Google Translate and DeepL. Two-level analysis revealed and classified errors of the technological layer (lexicographic insufficiency, transliterations, literalisms) and the cognitive-linguistic layer (loss of cultural connotations, incorrect communicative strategy). Quantitative and qualitative analysis demonstrated that for the majority of the examined units the systems produce either uncommented transliterations or semantically inadequate equivalents. It was established that the post-editing strategy should vary depending on the text type and the communicative situation: for a restaurant menu, transliteration with a functional explanation is optimal; for a tourist announcement, extensive cultural adaptation; for dialectal idioms, replacement with a functional analogue. The proposed model enables the systematic development of both machine translation literacy and the ability to perform functional and pragmatic text adaptation, meeting the demands of the translation services market and international standards.
This article examines the stylistic device of antiphrasis in modern online communication. Antiphrasis is a model of ambiguity in which a word, phrase, or sentence is used in a meaning opposite to its literal meaning, disguising a negative assessment with positive or neutral words. The paper emphasizes that antiphrasis is an effective tool for indirect speech influence, as it activates the addressee’s interpretive activity, forcing them to “complete” the hidden meaning. This study aims to identify and examine in detail the specifics of antiphrasis use in the social media. Particular attention is paid to analyzing its functional purpose and the variability of its implementation within the digital communication space. The relevance of this research stems from the extensive penetration of social platforms as a dominant channel of modern communication. This, in turn, stimulates the continuous evolution of language and generates innovative modes of linguistic expression, notably the intensified use of stylistic devices. A comprehensive understanding of antiphrasis patterns functioning in online discourse is the key to the implementation of an accurate analysis of current linguistic trends. The study’s findings are illustrated with examples from social media posts. The paper analyzes how antiphrasis can be combined with other speech techniques to enhance its impact, as well as how it functions in post titles, creating a certain expectation in the reader. Particular attention is paid to the multifaceted nature of the social media context. The conclusion emphasizes that antiphrasis in the modern digital space acts as an active, multifunctional tool of communicative influence, playing a significant role in shaping public opinion and criticizing social phenomena.
This study addresses the contradiction between the accessibility of online communication and the gradual simplification of national languages under the influence of digital mediatization processes. The relevance of the work stems from the need to overcome barriers to understanding meaning and protect the uniqueness of minority speech cultures from the algorithmic pressure of major network platforms. The empirical material includes the Russian-speaking communities of the Fediverse network, the decentralized blogging platform Bastyon, the secure Nostr protocol, as well as data from the Common Voice and Masakhane projects. Using texts in Russian, English, and African languages as examples, the paper thoroughly examines moderation rules, information distribution methods, and user speech behavior. The analysis confirms the negative impact of filter bubbles, which isolate audiences within homogeneous content and lead to the gradual disappearance of minority languages. It is established that the transition from centralized systems to decentralized environments eliminates algorithmic discourse distortion at the source code level. It is determined that eliminating automatic information ranking and transferring moderation to the interface level restores the subject’s speech agency, shifting communication behavior from passive to conscious. Based on the findings, the concept of an egalitarian digital discourse is substantiated, proving that ensuring the vitality of national languages requires rejecting their passive preservation in archives. The architectural reshaping of digital environments, which guarantees equal search visibility of content, genre freedom, and strict verbal responsibility of the author, is recognized as the only way out of the media communication crisis.
This article, using the digital linguistic paradigm, examines the temporal and aspectual factors that determine the expression of the taxis relations of simultaneity and multitemporality in modern German. The relevance of this study stems from the need to formalize linguistic phenomena for their subsequent automatic processing, as the category of taxis has long remained a “weak point” of NLP systems due to the complex interaction of lexical, morphological, syntactic, and discursive factors. The transition from an intuitive description of taxis meanings to formalized parametric models is a prerequisite for the creation of algorithms for the automatic determination of the temporal relationships of events in text. Using utterances with prepositional deverbatives (a total sample size of 15,622 contexts extracted from the Leipzig National Corpus), we analyze the interaction of four groups of parameters: the aspectual characteristics of the deverbative and finite verb (ultimate/inultimate, duration/momentariness), the semantics of the preposition (temporal or adverbial: modal, medial, conditional, causal), and the type of utterance (logically unconditioned, logically conditioned, multiple). We propose a formalized parametric model of taxis relations, presented as a tuple of features S = ⟨D, V, P, T, R⟩ and a decision table containing six mutually exclusive rules of the type “IF (D, V, P, T), THEN R”. The model is verified using corpus material with the participation of three independent annotators (Fleiss’s agreement coefficient κ = 0.87). It is argued that the key parameter differentiating the types of taxis relations is the opposition of liminality and inliminality, which integrates the temporal and aspectual characteristics of predicative units. It is shown that the temporal forms of the verb do not affect the relative temporal correlation, and prepositions with polycentric semantics (bei, mit, in, unter) generate syncretic adverbial-taxis relations. The results can be operationalized for subsequent automatic text processing, in particular, for the creation of algorithms for extracting taxis relations and annotated German corpora.
The article deals with a corpus-driven study of the lexeme security and its relevant word forms in English-language news texts aimed at developing analytical criteria that based on frequency characteristics reveal its semantic and pragmatic functions in the media discourse. The relevance of this research stems from the need for a comprehensive analysis of the concept security, identification of its functioning in contemporary media texts, and assessment of the pragmatic potential of the lexeme security in the dynamically changing information environment. The corpus approach is of particular importance as it enables the study of lexical items on the basis of a representative text collection, frequency patterns and collocations detection and specifics of contextual realization. The empirical material comprises an original balanced synchronic linguistic corpus of current CNN news texts for the period January 2025 – February 2026 of 120,942 sentences or over 3 million tokens totally. From this corpus 1,536 sentences containing 1,629 hits for the lexeme security were extracted. The technical platform for the study is the original software suite “Balanced Linguistic Corpus Generator and Corpus Manager.” The following methods were applied: exhaustive sampling, automatic natural language processing, including a stemming-based procedure, quantitative, collocational, semantic and pragmatic analyses, the original professionally oriented programming method, as well as elements of discourse analysis and qualitative content analysis. A comprehensive sample was compiled in which all relevant word forms of the lexeme security were identified; quantitative and qualitative analyses revealed frequent collocations and the principal directions of the lexeme functioning in the media text corpus. The notion of a research subcorpus was defined and ten interrelated corpus-oriented criteria were proposed to enable primary representation of the lexeme in the corpus, systematic description of its contextual realization and pragmatic interpretation, thereby providing an integrated corpus-driven approach from frequency to semantics.
The paper aims to establish the distribution of the functional load (FL) among the meanings of the English lexeme beauty across three chronological slices and to trace the dynamics of semantic temperature and entropy. The data sources include the Thorndike-Lorge Semantic Count (early 20th century, 1,000 contexts), the British National Corpus (1990s, 300 contexts), and the Corpus of Contemporary American English (2000s–2020s, 300 contexts). FL is defined as the proportion of a word’s occurrences in a given meaning relative to its total number of uses. BNC and COCA contexts were manually annotated for 7 meanings based on the OED; the distribution was approximated by the exponential model , whose parameters are interpreted by analogy with the Boltzmann distribution in statistical thermodynamics. The study reveals that semantic temperature T = 1/α shows a steady increase across slices (1.21 → 1.70 → 2.09), indicating a levelling of the semantic hierarchy and a decline in the dominance of the primary meaning. Substantive shifts were identified: a drop in the meaning “a beautiful woman” and a rise in the pragmatized meaning “advantage” (that’s the beauty of it) in American English compared to British English. The paper concludes that the FL method is applicable to the diachronic analysis of individual lexemes and can be extended to other languages.
This paper examines the phenomenon of native advertising in travel blogs on the Dzen platform through the lens of narrative coherence – the ability of an advertising message to maintain internal consistency, logical flow, and naturalness within the author’s narrative. The methodological framework includes the descriptive and analytical method (to identify recurring integration patterns), integrated into algorithmic procedures for identifying advertising content based on semantic and syntactic analysis in the GATE environment using ANNIE Gazetteer, JAPE Transducer, and Java Regexp Annotator technologies, the typologization method (to develop a typology of native integration tools), content analysis of user comments, narrative analysis (to assess story coherence and consistency), as well as the comparative method. Based on an analysis of 114 partner publications from two popular blogs (“TrueStory Travel” and “Travels with a Camera”), three levels of advertising integration are identified: narrative strategies (personal experience, problem solving, testing, myth debunking), discursive formats (review, recommendation, roundup, repost), and commercial markers (links, promo codes, disclosure labels). Analysis of user comments allows for a typology of audience reactions (approval, neutral discussion, skepticism, recognition of advertising, aggressive rejection) and establishes that reviews and roundups generate the least negativity, whereas the “myth debunking” tool and abrupt transitions to the product disrupt narrative coherence and provoke criticism. Principles for maintaining coherence are formulated: contextual relevance, genre appropriateness, the authorial figure as a guarantor of credibility, emotional engagement, and structural placement of direct calls-to-action at the end of the text. It is concluded that the effectiveness of native advertising on Dzen depends not only on camouflaging advertising content as editorial content but also on the quality of the advertising message’s embeddedness within the narrative.
This study is devoted to a corpus-based analysis of the literary space representation in J. R. R. Tolkien’s novella “The Hobbit, or There and Back Again”. The relevance of the research is due to the need to refine the theoretical foundations for describing locative categories in literary texts and to further develop corpus linguistic methods for the analysis of literary works. The material of the study is the original text of the novella organized into a balanced linguistic corpus comprising 5,648 contexts (sentences) or 113,650 tokens. The lexical units “mountain”, “river”, and “lake” were selected as the objects of analysis, representing the basic types of locative organization in the fictional world. The study employs methods of corpus analysis, contextual and semantic analysis, as well as descriptive analysis. The results show that the selected lexical units exhibit stable locative characteristics and perform a range of functions, including locative, delimiting, dynamic, axiological, symbolic, and structure-forming functions. It is demonstrated that “mountain” constructs a vertically organized locative model and functions as a center, boundary, teleological goal, and symbolic space of power and myth; “river” represents a linear-dynamic space of communication, movement, and boundary-making; “lake” realizes a model of enclosed, deep, and ambivalent space. It is concluded that the literary space in Tolkien’s work has an independent semantic organization and can be effectively described using corpus-based methods without necessarily invoking the chronotope category. The findings expand the potential of corpus linguistics in the study of locative categories in literary texts.
The study is devoted to a comparative analysis of the linguistic features of audio descriptions and art‑historical descriptions of art works. The relevance of the work stems from the need to develop effective methods for conveying visual information through verbal description for people with visual impairments. The research involved the analysis of a text corpus totalling 5 956 tokens, which includes descriptions of the painting “Summer” by N. Bogdanov‑Belsky and the sculpture “Mother and Child” by P. Trubetskoy. The technical platform utilizes the original software package “Balanced Linguistic Corpus Generator and Corpus Manager”. The research methodology included qualitative linguistic analysis (lexical, syntactic, stylistic, and pragmatic), as well as quantitative and statistical approaches. The analysis results revealed significant differences in the morphosyntactic characteristics of the texts. Audio descriptions are characterised by a predominance of concrete vocabulary, simple syntactic constructions, and a clear logical sequence of presentation. Art‑historical descriptions, in turn, are distinguished by a high level of terminological density, complex syntactic structures, and an academic style of exposition. Stylistic analysis has shown that both text types combine various linguistic means. However, audio descriptions aim to create the most accurate visual images through accessible sensory analogues, whereas art‑historical descriptions focus on expert interpretation of the works. Pragmatic analysis has revealed differences in target audience and communicative goals: audio descriptions are aimed at sensory compensation, while art‑historical descriptions are geared towards professional interpretation. At the same time, both text types demonstrate a high density of substantive vocabulary and a large number of adjectives. The practical significance of the study lies in the potential to use the obtained data for training audio description specialists, adapting art‑historical texts, and developing methodological materials in the field of audio description.
In order to identify the reflection of book traditions in seven Slavic miscellanies of the 11th–13th centuries — the Codex Suprasliensis, the Miscellanies of 1073 and 1076, the Uspensky, Troitsky, and Tolstovsky miscellanies, and the Miscellanies of the 13th century — a statistical analysis was carried out on nineteen pairs of synonymous words serving as markers of the Cyrillo‑Methodian, Preslav, and East Slavic scribal schools. To compare the widely varying token frequencies of lexemes across the miscellanies, a normalized weighted frequency method was used. This method consists of computing the term frequency (TF) of a word in a document, applying inverse collection term frequency (ICTF) weighting, normalizing the resulting TF×ICTF product by the maximum value of that word across all miscellanies, and then summing and averaging the normalized indices for words belonging to each tradition. The resulting value was interpreted as the tradition index for the given miscellany. Heat maps, dendrograms, and a principal component plot (PCA plot) were constructed, which revealed the statistical weight of the group of lexemes from each of the book schools within the miscellanies, as well as the similarity and contrast among the miscellanies and their parts. An interpretation of the statistical space of the miscellanies is proposed, taking into account information about them obtained through traditional methods of textual criticism, historical linguistics, and linguistically oriented textual analysis. It is shown that the statistical characteristics not only correspond to the existing knowledge in traditional scholarship regarding the language of the miscellanies and their parts, but also refine and supplement that knowledge.
The paper presents a diachronic analysis of inaugural speeches delivered by the leaders of the United States, the United Kingdom, and Russia over the period 2000–2025, aiming to identify the evolution of lexical dominants in political discourse under the influence of global socio-political transformations. The relevance of the study is determined by the need to understand changes in the verbalization of state identity and speech strategies in the context of 21st-century global crises, including economic instability, the COVID-19 pandemic, and increasing international tensions. Political discourse is considered as a tool for constructing social reality and shaping collective values in conditions of global turbulence. The research material comprises a corpus of 20 inaugural speeches by leading political figures of the three countries, selected according to genre specificity, temporal framework, and political-discursive significance. The main method applied is corpus analysis using Voyant Tools, which enabled the identification of frequent lexical units, their dynamics, and contextual relationships. Additionally, semantic grouping methods and qualitative interpretation within the framework of critical discourse analysis were employed. The results demonstrate the presence of stable national models of political rhetoric alongside a convergence of discursive strategies during crisis periods. It is established that American discourse retains an ideological orientation, British discourse is characterized by pragmatism and institutional focus, while Russian discourse demonstrates a state-centered and solidaristic model. Crisis events contribute to a shift from ideological concepts toward pragmatism, sustainability, and collective unity. The paper concludes that inaugural speeches function not only as reflections of political reality but also as active instruments of its construction, capturing transformations in cognitive models of power in the context of global challenges and an emerging multipolar world.
This article explores the polycode nature of film texts, which underlies two types of pragmatic information perception: cognitive resonance as an enhancement of emotional impact resulting from the semantic correlation of the film’s verbal, visual, and auditory components, and cognitive dissonance as a weakening or problematization of the pragmatic effect caused by a conflict between semiotic codes. To verify qualitative observations, the study employs digital linguistics methods: automatic audio transcription (whisper model), audio parameter analysis (RMS loudness, spectral centroid) using the librosa library, and topic modeling of dialogues (LDA). These methods allow quantifying the effects of cognitive resonance and dissonance and presenting the results in a measurable, reproducible form. Using documentary films about the blind – the Russian “City of the Blind” (2009), the British “Notes on Blindness” (2016), and the feature film “Deadpool” (2018) – this article analyzes scenarios of resonant and dissonant interactions between semiotic codes. It demonstrates that the polycode nature is a key mechanism for conveying “foreign” experience in contemporary screen narratives.
The aim of the study is to develop and test a corpus-based model for reconstructing the cognitive-discursive matrix of teacher professional identity. Professional identity is interpreted as a discursively represented configuration of linguistic, conceptual, terminological, pragma-axiological and metaphorical parameters. The empirical basis of the study is a specialised corpus of English-language professional pedagogical discourse comprising more than 2,700 pages and over 2 million word tokens. The corpus includes written and oral materials: monographs, textbooks, research and popular-science articles, lectures, and professional educational texts. The research methodology combines digital text preprocessing, frequency analysis, reconstruction of dominant concepts, analysis of terminological density, pragma-axiological modelling, and identification of conceptual metaphors. The study established that the core of the teacher’s cognitive-discursive matrix is formed by six dominant concepts: “education”, “language”, “culture”, “identity”, “research”, and “method”. An inverse functional relationship was identified between terminological density and pragma-axiological load: “method” demonstrates the highest degree of professional specialisation and the lowest axiological load, whereas “identity” and “culture” display the strongest value-related charge. The analysis of conceptual metaphors revealed 13 dominant models, among which “education is a journey”, “learning is growth”, “teacher is a guide”, and “teaching is growth” constitute the core models. The study concludes that corpus-based modelling makes it possible to transfer teacher professional identity from the domain of interpretative description to the plane of reproducible analysis of discourse parameters, while the cognitive-discursive matrix serves as an effective operational tool for reconstructing teacher professional subjectivity within the digital linguistic paradigm.
The present study addresses naming strategies in different types of digital environments, an aspect that has not yet been sufficiently and systematically explored within digital onomastics. While existing works focus predominantly on the semantics and classification of nicknames, this article investigates how the degree of environmental regulation affects the functional relationship between a nickname and the image of its bearer. The subject of the study is the mechanisms that shape either an immanent or a situational connection between a name and the bearer’s image depending on the closedness or openness of the communicative space. The empirical material consists of game nicknames from the forum role-playing games Legends of Empia (397 units) and Spirits of the Earth (80 units), which exemplify a regulated environment, as well as user nicknames from the social network X (111 units) and 161 thread reactions, which exemplify an open discursive environment. The aim of the study is to reveal how the degree of discursive regulation determines the differences in naming strategies in closed and open digital environments and, on this basis, to construct an explanatory model. The analysis has established that within the closed discourse of forum role-playing games, the anthroponym functions as a poetonym with an immanent connection to the image and a narrow field of interpretation determined by the rules of the game world. In the open environment of the social network “X” (a social network blocked in the Russian Federation), a situational-performative strategy is realized: the meaning of the name is not fixed in advance but is formed in the act of communication under the influence of extralinguistic factors, which generates a multiplicity and often conflictuality of interpretations. The proposed model makes it possible to systematize naming strategies depending on the type of digital environment. It can be applied to the analysis of nicknames in various online communities – from gaming platforms to professional forums – thus contributing to the development of digital onomastics as a branch of the digital humanities.
This study examines the use of currency denominations in a balanced linguistic corpus of news texts from “Focus”, Germany’s largest magazine. The relevance of the work is supported by the high interest in news discourse analysis within the linguistic community, as well as the objective need to develop mechanisms for identifying hidden manipulative influence on mass audiences through the media. The study textual material was transformed into an SQL database with a volume of over 1.7 million tokens using the Russian original software package “Balanced Linguistic Corpus Generator and Corpus Manager” (Rospatent Certificate No. 2023683209). The research utilized methods of automatic data collection, analysis of global currency data, corpus queries, quantitative analysis, and elements of discourse analysis (interpretation of contexts based on data from a broad discursive field). As a result, we obtained a subcorpus of 2,735 contexts containing the usage of currency names, compiled a frequency ranking, and conducted an in-depth qualitative analysis of the most common topics. Among these, we identified the state of the German national economy as a dominant theme, including comparisons and contrasts with the economies of other EU countries and Russia. We also noted the active use of currency names in creating highly connotative contexts for the ideological component associated with denigrating Russia in connection with the Special Military Operation and creating a negative image of Iran in connection with US military aggression against that country.
This study explores developing inclusion competence in future intercultural communication specialists by integrating work with works of art that present the theme of disability into the educational process. The relevance of this work stems from changes in contemporary Russian society, the digital transformation of education, and the search for one’s own identity, including the desire to understand certain key concepts within Russian tradition and culture. One concept requiring such understanding is disability. An analysis of the foreign language syllabi at Moscow State Linguistic University revealed that the topic of disability is not represented in educational materials. The authors propose integrating this topic into existing sections of the “Art” and “Culture” curricula. Using a comparative analysis of literary representations, the study describes the differences between Russian and Western models of understanding inclusion. The authors developed criteria for selecting works: a combinatorial principle, authentic experience, external reflection, consideration of the historical evolution of various models of disability perception, as well as the linguadidactic and discussion potential of the materials. The paper also proposes and describes a methodology for working with literary texts, which includes tasks to master ethical vocabulary, develop clear language skills, and utilize discourse analysis. To address new linguadidactic challenges in the context of digitalization of education, the authors substantiate the use of digital tools (Lingro, AnsewerGarden, WAVE, StoryBird, LMS Moodle) to create interactive dictionaries and tag clouds, assess the accessibility of the digital environment, simulate real-world experiences, and conduct effective project work. The authors conclude that integrating disability through art into foreign language courses during the training of intercultural communication specialists not only fosters inclusive competence as a component of professional training but also enriches students’ linguistic and cultural horizons. The proposed materials can be adapted for a wide range of educational programs in the humanities.
