Research

The research of the Multilinguistics project is organised in independent fronts, each devoted to a delimited aspect of the linguistic transformations associated with the digital era. The fronts draw on the project's disciplinary domains as lenses of analysis and may combine several of them in a single study. The glossary defines these and other terms.

Domains of analysis

1. Structural linguistics: sounds, forms, syntax, meaning and use.
2. Historical linguistics: change, ancestry, divergence, convergence and contact between languages.
3. Cognitive linguistics: acquisition, representation, processing and multilingual cognition.
4. Sociocultural linguistics: communities, multilingual practices, variation, identity and power.
5. Mathematical and computational linguistics: formal models, corpora, statistics and language technologies.
6. Epistemology of linguistics: explanatory assumptions, universals and the limits of inference.

A seventh domain, of integration, brings together the results obtained through the different lenses.

Research fronts

Digital multilingualism active

Studies language contact, variation, the informational dimension and the computational modelling of Spanish, Portuguese and English in digital environments. It hosts the project's first study, a scoping review of the literature, and has a related publication. Front page.

Generative artificial intelligence and written style active

Investigates whether and how writing assisted by generative artificial intelligence changes written style, from the individual text to the norms of a community: differences between texts produced with and without assistance, convergence between writers who use the same tools, the persistence of effects after assistance is withdrawn, and the possibility of long-term collective change. A mathematical model of human writing is under development, in which the author is represented as a stochastic operator that transforms a tensor representation of the task into a tensor representation of the text; the model will serve as the basis for comparing writing with and without the assistance of language models.

Regional academic writing planning

Examines, in the diachronic corpus of the Revista Científica Humanidades of the Universidad Autónoma Juan Misael Saracho (Tarija, Bolivia; ISSN 2707-4072 and 2789-4789), formed by the thirteen issues published between 2018 and 2026, how digital transformations are reflected in academic writing in Spanish: citation practices, lexical, syntactic and register features, and possible traces of writing assisted by artificial intelligence.

Language teaching and digital technology planning

Investigates the use of digital tools and generative artificial intelligence in the teaching and learning of Spanish, Portuguese and English, and the contributions that the project's results can make to the design of teaching materials and to assessment.

Machine translation with large language models planning

Investigates translation between Spanish, Portuguese and English carried out by large language models: how it differs from human translation, how it treats regional varieties and registers, the possible influence of machine-translated texts on usage and norms in the target languages, and the methods suited to evaluating it.

Indigenous languages in digital environments future stage

A theme planned for a later stage: the presence, use and contact of the Indigenous languages of the Andes and the Southern Cone with Spanish and Portuguese in digital environments.

The research questions of the fronts in planning are being defined. Researchers interested in contributing to any of them may consult the Collaborate page.