Call for 2027 dossier

Linguistics in AI and AI in Linguistics

Guest Editors
    • Raquel Freitag – Federal University of Sergipe
    • Livy Real – Federal University of Amazonas

This dossier departs from a dual premise: linguistic research can and should contribute to the development of more robust, inclusive, and theoretically grounded AI systems; and AI systems, by operating with language, raise questions that Linguistics is uniquely positioned to answer – about variation and change, about prosody and intonation, about pragmatics and speech acts, about what constitutes a conversational turn, and about what it means to “understand” a text. The dialogue between these two fronts is both urgent and still incipient.

We invite researchers to submit papers exploring this intersection from theoretical, methodological, and applied perspectives, including, among others, the following thematic lines:

Linguistic data and model training: Collection, documentation, annotation, and curation of corpora for AI systems; Representation of Brazilian linguistic diversity (regional, social, age-related, and register variation), and other countries linguistic diversity; Risks of bias in models trained on non-representative data; Best practices in Open Science applied to linguistic resources.

Linguistic theory and system architecture: What Phonology, Morphology, Syntax, Semantics, Pragmatics, and Conversation Analysis (and beyond) have to say about the functioning and limitations of LLMs, TTS, STT, and machine translation systems; Implicit linguistic assumptions in computational architectures; The concept of token and its implications for a theory of language.

Applications and linguistic evaluation of systems: Linguistic quality of AI outputs (coherence, pragmatic adequacy, register, hallucination, representation of varieties); Evaluation of speech synthesis and recognition systems regarding prosodic naturalness and representation of non-hegemonic varieties; Prompt engineering as a practice of applied pragmatics.

Activist Linguistics and data policies: Linguists as agents in the public debate on AI; Participation of the linguistic community in data policy formulation, especially within the context of the Brazilian Artificial Intelligence Plan; Challenges for building linguistic data infrastructures in Brazil; Equity in access to language technologies.

Ethics, transparency, and social impacts: Reproduction of linguistic prejudices by AI systems; Accessibility of language systems for populations with minority linguistic varieties; Voice cloning and identity; Authorship, transparency, and integrity in AI-mediated writing; Impacts on language education; 
Empirical, theoretical, and critical literature review contributions are welcome, in Portuguese, English, Spanish, or French.

Schedule:

• Until June 30, 2026: send a 100-250 word abstract about the proposed work to the email: rkofreitag@uol.com.br.

• Until August 1, 2026: authors of the abstracts will receive feedback on the relevance of their proposal to the call for papers.

• Until October 31, 2026: submit the complete article through the Texto Livre journal's submission system, following the guidelines available at: https://periodicos.ufmg.br/index.php/textolivre/about/submissions

• From March 2027 onwards: publication of the dossier.

We strongly encourage prior deposit in preprint repositories, such as arXiv (https://arxiv.org), or SciELO Preprints (https://preprints.scielo.org) providing the access identifier (preprint DOI) in the Open Science Compliance Form available in the journal system.