Minority and Minoritized Languages in Artificial Intelligence: Visibility, Representation, and Inclusion
Langues minoritaires et minorisées dans l’intelligence artificielle : visibilité, représentation et inclusion
“Ikhtilaf” Journal of Critical Humanities and Social Studies
« Ikhtilaf », revue de sciences humaines et sociales critiques
Published on Wednesday, May 13, 2026
Abstract
As Artificial intelligence (AI) reshapes how language is processed, generated, and transmitted, a critical disparity is emerging : the vast majority of natural language processing (NLP) systems, large language models (LLMs), and machine translation tools are built around a handful of dominant languages — leaving thousands of languages with little or no digital presence. This asymmetry raises urgent questions about linguistic justice, cultural preservation, and epistemic equity. This special issue of Ikhtilaf, Journal of Critical Humanities and Social Studies seeks to critically examine these dynamics, foster interdisciplinary dialogue, and amplify voices from communities whose languages are at risk of digital — and ultimately cultural — erasure.
Announcement
About the Journal
Ikhtilaf — Journal of Critical Humanities and Social Studies is a peer-reviewed academic publication hosted by the Moroccan Institute of Scientific and Technical Information (IMIST). The journal is committed to fostering critical intellectual debate across the humanities and social sciences, with a particular focus on issues of language, culture, society, and knowledge in Moroccan, North African, and global contexts.
Argument
Ikhtilaf — Journal of Critical Humanities and Social Studies invites scholars, researchers, and practitioners to submit original contributions for a special issue dedicated to the place of languages left at the margins of artificial intelligence systems — those rendered invisible by technological, colonial, or institutional forces, and those for whom digital presence remains fragile, contested, or entirely absent.
As AI reshapes how language is processed, generated, and transmitted, a critical disparity is emerging : the vast majority of natural language processing (NLP) systems, large language models (LLMs), and machine translation tools are built around a handful of dominant languages — leaving thousands of languages with little or no digital presence. This asymmetry raises urgent questions about linguistic justice, cultural preservation, and epistemic equity.
This special issue seeks to critically examine these dynamics, foster interdisciplinary dialogue, and amplify voices from communities whose languages are at risk of digital — and ultimately cultural — erasure.
Themes and Topics of Interest
Submissions are welcome on topics including, but not limited to :
- Language Technology & Low-Resource NLP
- Speech recognition, machine translation, and text-to-speech systems for low-resource languages
- Challenges and innovations in training large language models (LLMs) on low-resource linguistic data
- Few-shot and zero-shot learning approaches for minority language NLP tasks
- Morphological analysis, POS tagging, and syntactic parsing for morphologically rich languages
- Evaluation frameworks and benchmarks adapted to low-resource language contexts
- Data Preparation : Translation, Standardisation & Corpus Building
- Translation as a bootstrapping strategy : building parallel corpora for low-resource languages
- Language standardisation as a prerequisite for NLP : orthographic, morphological, and lexical normalisation
- Designing and curating bilingual and multilingual aligned corpora
- Human translation workflows and quality control for training data production
- Cross-lingual transfer learning enabled by translated parallel texts
- The role of institutional actors (governments, academies, universities) in data standardisation
- Community participation in data collection, validation, and annotation
- Ethical and political dimensions of standardisation : whose variety of the language gets encoded ?
- The Changing Web : Memes, Multimodal Data & AI-Generated Content
- From web writing to memes and short-form content : what the shift in digital communication means for low-resource language NLP
- The minoritized language web : mapping the actual digital presence of marginalized languages online
- Memes, emojis, and code-switching as linguistic data : challenges and opportunities for NLP
- The contamination problem : how AI-generated content in dominant languages crowds out authentic minority language data on the web
- Voice messages, audio content, and informal digital registers as untapped corpora
- New methods for collecting, filtering, and curating non-standard web data for low-resource language AI
- Social media, messaging platforms, and community archives as alternative data sources
- The invisibility of oral and multimodal language practices in current NLP paradigms
- Cultural Preservation & Digital Revitalization
- AI-assisted documentation and archiving of oral traditions, dialects, and endangered languages
- Digital tools for language education and intergenerational transmission
- Community-driven approaches to AI development for language revitalization
- Case studies from Amazigh, Darija, African, Indigenous, and other minority language communities
- Ethics, Equity & Linguistic Justice
- Algorithmic bias and the underrepresentation of minority languages in AI systems
- Power dynamics, language hierarchies, and the politics of AI language design
- Decolonial approaches to natural language processing and AI development
- Policy frameworks for inclusive and equitable AI language governance
- The role of multilingualism in AI fairness and accessibility
- Sociolinguistics & Critical Theory
- Language endangerment in the digital age : new threats and opportunities
- Identity, belonging, and the digital exclusion of language communities
- Translanguaging, code-switching, and hybrid linguistic practices in AI contexts
- Interdisciplinary perspectives from linguistics, anthropology, education, and media studies
Submission Guidelines
- Manuscripts must be original, unpublished, and not under review elsewhere.
- Papers may be submitted in English, French, or Arabic.
- Manuscripts should be between 6,000 and 10,000 words, including references.
- All submissions must follow APA 7th edition referencing style.
- Abstracts of 200–250 words and 5 keywords must accompany each submission.
- All manuscripts are subject to double-blind peer review.
- For submission guidelines report to the journal website.
Submissions should be made through the journal’s online platform.
Important Dates
- Submission Deadline: 30/09/2026
- Publication Date : January 2027
For inquiries regarding this Call for Papers, please contact :ikhtilaf@gmail.com
Guest editors
• Touaf Larbi — Université Mohammed Premier, Oujda (Maroc)
• El Jarari Mohamed — Université Mohammed Premier, Oujda (Maroc)
Subjects
- Language (Main category)
- Mind and language > Language > Linguistics
Date(s)
- Wednesday, September 30, 2026
Attached files
Keywords
- intelligence artificielle, ia
Information source
- Mohamed El jarari
courriel : mohamed [dot] jirar1981 [at] gmail [dot] com
License
This announcement is licensed under the terms of Creative Commons CC0 1.0 Universal.
To cite this announcement
« Minority and Minoritized Languages in Artificial Intelligence: Visibility, Representation, and Inclusion », Call for papers, Calenda, Published on Wednesday, May 13, 2026, https://doi.org/10.58079/167mh
