International Journal of Multilingualism and Languages for Specific purposes

Algorithmic Mediation and Linguistic Futures: Power, Pedagogy, and Practice in Multilingual Worlds

2026-03-06T19:55:27+00:00

This issue of the International Journal of Multilingualism and Languages for Specific Purposes (IJMLSP) brings together cutting-edge research that interrogates the evolving relationships among language, technology, policy, and society in an era increasingly shaped by artificial intelligence and intensified multilingual contact. Across diverse geopolitical contexts, ranging from maritime academies and secondary schools to legal frameworks governing indigenous languages, the contributions collectively illuminate how linguistic practices are being reconfigured through algorithmic mediation, institutional regulation, and sociocultural negotiation. Several articles foreground the pedagogical implications of AI-driven tools in English for specific purposes (ESP), bilingual education, and professional training. These studies demonstrate that artificial intelligence, when deployed critically and context-sensitively, can enhance learner engagement, professional simulation, and curriculum innovation, while simultaneously raising concerns about overreliance, epistemic authority, and disciplinary misalignment. Complementing these applied perspectives, other contributions adopt linguistically grounded and sociopolitical lenses, examining morphosyntactic flexibility in Yoruba, computational marginalization in low-resource languages such as Malayalam, and code-switching as identity work in African and South Asian higher education contexts. A central unifying thread across the issue is the tension between inclusion and control. Whether manifested in language policy discourse, AI-mediated assessment, or institutional language choice, linguistic inclusion often coexists with subtle mechanisms of regulation and hierarchy. The critical discourse analysis of Taiwan’s Indigenous Language Development Act exemplifies how legal language symbolically recognizes rights while delimiting their practical enactment, a dynamic echoed in educational and technological settings elsewhere in the volume. Taken together, the articles argue for a reconceptualization of multilingualism not merely as linguistic diversity, but as a dynamic field shaped by power, ideology, technology, and agency. This issue thus advances an interdisciplinary agenda that bridges applied linguistics, sociolinguistics, computational linguistics, education, and critical policy studies, offering timely insights for scholars, educators, and policymakers committed to more equitable and sustainable linguistic futures.

Maritime English Instruction for Future Ship Engineers: A Pedagogical Study Using ChatGPT and Digital Learning Platforms

2026-03-06T19:55:25+00:00

The paper is devoted to the use of Artificial Intelligence while teaching Maritime English to future ship engineers. The analysis of scientific literature was done in the article. The research shows a limited number of works devoted to the use of Artificial Intelligence in maritime education and training. The experimental study was conducted at the Ukrainian Maritime Academy and college. Diagnostic testing, questionnaires, various interactive teaching methods, professionally oriented tasks, digital tools, and instruments were used during the experiment. Moreover, the pedagogical experiment was conducted to test the effectiveness of motivational assessments, interactive and professionally oriented methods of teaching Maritime English in the training of future ship engineers. The sample consisted of 153 cadets from the Marine Engineering Faculty. They were divided into experimental and control groups. Quastionnaires identified the dynamics of cadets’ motivation in the AI use while e-learning. The DataIsland platform with integrated ChatGPT was chosen as the main Artificial Intelligence tool. LMS Moodle was used as the main platform to contain interactive tasks, quizzes, etc. Systematic processing and analysis were carried out to verify the validity of the results. Quantitative and qualitative approaches were combined to ensure objectivity of the experiment. Statistical analysis was used to process the data. Cadets studied the following topics in Maritime English: engine room equipment, its maintenance, instructions on emergencies, watchkeeping procedure, etc. The integration of artificial intelligence into the learning process allowed the use of ChatGPT to simulate dialogues in professional scenarios. ChatGPT was also effective in creating case scenarios. Another advantage of ChatGPT is its effectiveness in creating glossaries and creating interactive exercises. According to the results, the experimental group achieved greater progress in learning Maritime English, which is very important for ship engineers. The results also prove that AI in Maritime English e-learning should be used as a supportive and motivational tool. This research has important practical implications for teachers (e.g., design of professionally oriented tasks, teachers’ digital competence development).

AI-Powered Writing Assistance Tools in the Foreign Language Classroom: Unveiling ESP Teachers’ Perspectives across Disciplines

2026-03-06T19:55:24+00:00

This mixed-methods study examines Vietnamese English for Specific Purposes (ESP) teachers’ perceptions of AI-powered writing assistance tools and explores how these perceptions differ across academic disciplines. Data were collected from 40 ESP teachers through a questionnaire and semi-structured interviews with 15 participants. Descriptive statistics and thematic analysis were employed to investigate teachers’ attitudes, perceived benefits and challenges, and contextual factors influencing AI integration. Quantitative findings indicate strong agreement on the pedagogical value of AI tools in promoting learner autonomy (M = 4.00, SD = 0.96), enhancing engagement (M = 3.78, SD = 0.91), and supporting collaborative learning (M = 3.82, SD = 0.89). However, teachers expressed significant concerns regarding the accuracy of AI-generated language (M = 4.82, SD = 0.90), students’ overreliance, and difficulties aligning AI tools with existing curricula. Qualitative data revealed notable disciplinary differences: teachers in hard sciences highlighted limitations in domain-specific terminology and precision, whereas those in soft sciences perceived stronger benefits for general academic writing and motivation. The findings underscore the need for discipline-sensitive implementation, clearer institutional guidelines, and targeted professional development to support responsible and effective AI integration in ESP contexts. This study contributes to the emerging literature on AI in language education and offers practical implications for policymakers, curriculum designers, and educators seeking to optimize AI-assisted writing instruction.

AI-Driven Curriculum Innovation in Multilingual Secondary Education: Evidence from English and Yoruba Instruction in Ondo State, Nigeria

2026-03-06T19:55:22+00:00

The increasing linguistic diversity of Nigerian secondary schools continues to challenge the capacity of traditional curricula to address learners’ varied language needs, particularly in bilingual English–Yoruba contexts. Within this context, artificial intelligence (AI) offers new possibilities for curriculum innovation that can personalise instruction and support indigenous language education. The study employed the mixed-methods approach. The qualitative data were gathered through curriculum analysis, teacher interviews, and student focus groups to study the practice, perception, and contextual needs. The results of a quasi-experimental design were obtained by means of quantitative data, which were collected on a sample size of 120 Senior Secondary One (SSI) students in four schools. Students were taught with AI tools, namely:Grammarly and ChatGPT for English and YorubaBERT, Duolingo Yoruba, and chatbots for Yoruba, for six weeks of intervention. Proficiency was tested using pre- and post-tests, and surveys and logs were used to measure engagement and attitudes.Results showed that English and Yoruba curricula were strongly different, with the latter having less organization and being not tech-friendly. Positive attitudes toward the use of AI were reported by both students and teachers, but interactive feedback and cultural relevance were found to be the most beneficial features. The statistical findings demonstrated a significant improvement in both English and Yoruba, although English was gaining more because of the maturity of the tools. Among the challenges were poor ICT infrastructure, poor teacher training, and insufficient resources forAI in the country, whereas the opportunities were in the area of interaction, adaptive learning, and cultural preservation. The study concluded that the innovation of curriculum with the use of AI promoted bilingual learning by connecting the global technologies with the local sociolinguistic conditions. It recommended the investment in AI tools, training of teachers, as well as the infrastructure, with a view to facilitating inclusive and sustainable education.

Morphosyntactic Fusion Processes in Yoruba: A Systemic Functional Analysis of Fused-Head Constructions

2026-03-06T19:55:21+00:00

Yoruba morphosyntax has exhibited some level of complexity in the domains of fusion and headless construction, as these structures emerge from derivational processes with flexible and thin categorical boundaries. The study investigated fused-head constructions from a Systemic Functional Grammar perspective. The study adopts a corpus-based approach to examine the morphosyntactic behaviour of fused heads in Yoruba. Drawing on written data from the Yoruba Web Corpus (yoWaC) hosted on Sketch Engine, the analysis quantitatively examines the frequency and distribution of derived compound forms, with particular attention to their categorial outcomes (nominalized as nouns or denominalized as adjectives). The study focuses on compounds formed with the affixes a- and o-, illustrated by forms such as Olólà ‘noble/wealthy’ and Àrẹwà ‘beauty queen’, and identifies recurrent morphosyntactic patterns through systematic corpus querying and frequency analysis. Results show that affixed compound words are more frequently used as nouns than as adjectives or other de-nominalized forms in actual language use. More notably, some words outstandingly appear in their nominalized form with very few or no de-nominalized occurrences. The study posits that while Yoruba-affixed compounds tend to function more as nouns rather than adjectives, some words exhibit high flexibility, allowing them to function between nominalized and de-nominalized usages depending on the context. The findings help readers understand how the morphosyntactic flexibility of derived words in Yoruba affects headless constructions. The study also affirmed that the frequent use of post-modification through PRGs in nominalized affixed compounds can be interpreted as a way to avoid ambiguity and allow for precise categorization of the head noun. The high occurrence of nominalized heads suggests a productive nominalization system in Yoruba, where the affixes are used to create person-denoting nouns. This study offers broader theoretical insights into general morphosyntactic processes such as word formation, derivational processes, and category flexibility, thereby enriching cross-linguistic discussions of compound formation and morphosyntactic structure.

Challenges in Natural Language Processing for Malayalam: A Low-Resource Language Perspective

2026-03-06T19:55:19+00:00

Malayalam, a prominent Dravidian language chiefly spoken in the Indian state of Kerala, poses distinct difficulties for Natural Language Processing (NLP). Although it boasts over 35 million speakers, the language is still underrepresented in computational linguistics studies relative to high-resource languages like English or Hindi. This paper systematically examines the structural, sociolinguistic, and technological barriers to developing effective NLP systems for Malayalam, with particular attention to its agglutinative and highly inflectional morphology, diglossia and register variation, prevalent code-mixing with English, and non-standard Romanized writing practices known as Manglish. Through a theoretical and analytical approach, the study synthesizes and critically evaluates findings from Malayalam-specific and low-resource NLP research and assesses the applicability and limitations of computational methods such as multilingual pre-trained models, transliteration systems, and transfer learning techniques in the Malayalam context. The analysis finds that morphological complexity, limited annotated corpora, and inconsistent orthographic conventions significantly reduce model performance and limit the effectiveness of standard NLP pipelines. While multilingual and cross-lingual models partially address these issues by leveraging resource-rich languages, their effectiveness for Malayalam remains constrained by insufficient language-specific data and inadequate handling of morphological and register variation. This paper contributes by clearly identifying the core linguistic and resource-based challenges in Malayalam NLP and by proposing practical, research-informed strategies for corpus development, model adaptation, and hybrid linguistic-computational approaches. By situating Malayalam within broader discussions on low-resource and multilingual NLP, the study provides insights to inform future research and model development for Malayalam and other underrepresented languages, thereby promoting more inclusive and linguistically diverse digital technologies. The paper also highlights Malayalam's typological value as a morphologically rich Dravidian language, contending that knowledge from its analysis may guide the development of natural language processing (NLP) systems for structurally related languages. It also describes possible directions for future research, such as focused corpus growth, better annotation techniques, and low-resource evaluation systems. The report aims to close the gap between descriptive studies and workable solutions by outlining these directions.

Performing Inclusion, Enforcing Control: A Critical Discourse Analysis of Taiwan’s Indigenous Language Development Act

2026-03-06T19:55:17+00:00

Legal frameworks are often positioned as instruments for protecting and promoting Indigenous language rights. In Taiwan, the Indigenous Language Development Act (ILDA) is presented as a key policy for linguistic justice and revitalisation. However, how such rights are discursively constructed and governed through legal language remains under-examined. Drawing on van Leeuwen’s ‘construction of purpose’ framework within the Critical Discourse Studies approach, this study critically examines ILDA as a form of specialised legal and multilingual discourse. The analysis focuses on grammatical ambiguity, conditional modality, and the attribution of agency to examine how linguistic rights are framed and operationalised within the policy. The findings reveal that while the ILDA aims to protect and promote indigenous language rights, the study critiques the empowerment discourse of the policy as serving the dominant ideological interests. Through rigid structural and grammatical means, authority and ideology are structurally embedded within policy texts. Although the government positions itself as supportive with symbolic inclusion, the controlling nature of the policy mechanism limits the revitalisation efforts, thus functioning as a discursive mechanism of multilingual governance, legitimising state control while appearing to promote linguistic justice, discursively recognising linguistic rights but constraining them institutionally. By treating legal texts as specialised multilingual discourse, this study demonstrates that power and rights are mediated through rigid structural means and that multilingualism is carefully managed rather than supported. The study contributes to global discussions on the role of language policy in sustaining minority languages, suggesting that closer attention should be paid to how purpose and agency are constructed to inform more equitable policy design.

One Person, Many Tongues: A Domain-Based Evaluation of Multilingualism Among Igbo Students in a Nigerian University

2026-03-06T19:55:16+00:00

This study examines the linguistic experience of Igbo students in the University of Lagos who were proficient in four languages: two native languages (Igbo, Yoruba) and two foreign languages (English, French). This is with a view to ascertaining the factors affecting the choice of language in three different domains, such as home, classroom, and hostel room in school. Fishman’s Domain Analysis was used as a guide in the study. Through purposive sampling, one male and two female Igbo students studying Igbo and French, respectively, were selected. The limited number is attributable to the challenge in finding participants who met the criteria of proficiency in two indigenous and international languages. Based on the qualitative approach, a face-to-face unstructured interview was utilized to elicit information. The pooled data was validated through the member checking technique, while content analysis was used in data interpretation. It was found that language choice was mainly influenced by domains but also influenced by the needs of participants. English contested language space in the home with Igbo, including the intrusion of French, as well as Yoruba, the language of the host community. In contrast, indigenous languages also intruded into formal spaces in school, such as the hostel and classroom. Among factors that influenced language choice in the study were family language policy, visit of native language users, exogamous marriage, technology, and exclusion. A significant finding is that the four languages were not kept apart in use; rather, they were used jointly, with English being dominant in code-switching processes. Notably, multiple language use is adjudged beneficial to participants in this study as it promotes diversity and inclusion, which are needed in a multiethnic institution such as the University of Lagos, Nigeria. It is posited that a qualitative approach advances research in multilingualism given the importance of exploring the participants’ subjective opinion.

Switching Codes, Shaping Selves: Investigating Bangladeshi Undergraduates’ Linguistic Practices

2026-03-06T19:55:14+00:00

In Bangladesh, code-switching has become an increasingly natural and common part of communication among university students, reflecting their linguistic flexibility and social interaction patterns. Although the country is primarily monolingual, English has coexisted with Bangla since the British colonial era. However, empirical research examining students’ motivations and attitudes toward code-switching in public universities remains limited. This exploratory case study analyzes students' motivations and attitudes regarding code-switching at Jagannath University. A mixed-methods approach was employed, comprising a structured questionnaire (N = 25) and semi-structured interviews with 10 participants across five departments. Descriptive statistics were used to analyze survey responses, and thematic analysis was applied to interview data to identify recurring patterns and motivations. Quantitative results show that 80% of participants reported using code-switching to facilitate effective communication, and 36% cited lexical gaps as a key reason. Attitudinal data reveal that 60% of students perceived code-switching positively in academic contexts, while 36% expressed neutral views. Additionally, 68% of respondents believed that code-switching supports second or foreign language learning. Interview data further demonstrate that students employ code-switching to enhance clarity, express identity, and align with peer norms shaped by social media and popular culture. Attitudinal data indicate that students have a positive attitude toward code-switching, though some recognize potential difficulties in academic communication. Drawing on Myers-Scotton’s Markedness Model, Bangla emerges as the unmarked choice in informal interactions, whereas English functions as a marked choice to index prestige, competence, or modern identity. However, the findings are limited by the small sample size and focus on a single public university, which may constrain broader generalizability. Despite this limitation, the study offers empirical evidence from a Bangladeshi public university and highlights the need for longitudinal and comparative research. The study offers implications for language pedagogy in postcolonial contexts and advances our understanding of bilingual practices in higher education.

Students’ Cognitive Attitudes Towards African Languages Use in Lecturer Consultations at NWU–Potchefstroom Campus

2026-03-06T19:55:12+00:00

Language attitudes as a field of study can be divided into three categories: namely, cognitive, affective, and behavioural attitudes. The cognitive attitudes are related to learning, while the affective attitudes relate to emotions and feelings, and the behavioural attitudes have to do with action towards a language. In general, language attitudes of black South African students towards the use of African languages in Higher Education (HE) are conflicted. Studies have shown that students see the benefits of using African languages for teaching and learning, but still do not show very positive attitudes towards the adoption of these languages. In this study, mixed method approach was used, and I report on data from a broader PhD study. The research question that this paper answers is what the cognitive attitudes of students who use their languages during consultations are towards the use of those languages in HE. The data was collected through a general questionnaire to students who wished to participate, and another questionnaire for students who came to consult using their languages, followed by interviews and a focus group discussion for the latter. The cognitive attitudes of the participants reveal that students feel conflicted about the use of their languages in HE, despite being allowed to use them such as during consultations with their lecturer. The study found that students generally expressed a positive cognitive attitude towards the use of their languages in HE, citing understanding and overall expression and communication as the main benefits, but they also have conflicted cognitive attitudes that their languages can hinder understanding for others. Language awareness campaigns should be driven, which deal specifically with persuading the attitudes of students to be more positive towards the use of African languages in HE and their adoption, as this can cognitively benefit black South African students with understanding and communication.

Fading First Words: Early Childhood Education and the Shifting Place of Malay

2026-03-06T19:55:11+00:00

The Malay language is a fundamental pillar of Malaysia’s national identity, representing cultural unity, social cohesion, and shared heritage. In recent years, however, a noticeable decline in its use has raised serious concerns about the long-term sustainability of the national language. Comparable patterns of language shift have been observed in other multilingual societies, such as Singapore and Canada, where dominant global languages increasingly influence everyday communication. Early Childhood Education (ECE) is widely recognized as a critical stage for shaping linguistic identity and language practices from an early age. International evidence, particularly from New Zealand, demonstrates that ECE can function as an effective platform for language revitalization and maintenance. Despite these insights, limited scholarly attention has been paid to the role of ECE in sustaining the Malay language within the Malaysian context, especially from the perspectives of ECE educators and postgraduate students. This qualitative study aims to explore the perceptions of ECE educators and postgraduate students regarding the role of preschool education in maintaining Malay as the national language. It also seeks to identify key risk factors contributing to language shift among younger generations. Data were collected through in-depth interviews with nine participants from diverse backgrounds, including members of Malaysia’s three major ethnic groups, ECE postgraduate students, and teachers working in both public and private preschools. A thematic analysis approach was employed to identify recurring patterns and salient themes across participants’ accounts. The findings reveal a clear tension between efforts to preserve the Malay language and the increasing emphasis on early English exposure. Family language practices, institutional policies within preschools, and broader societal expectations were identified as influential factors shaping children’s language use. These findings underscore the need for balanced and context-sensitive ECE language policies that promote national language maintenance while acknowledging global linguistic demands. The study offers implications for policymakers, educators, and researchers, and provides insights relevant to other multilingual contexts facing similar challenges of language shift and preservation.