Abstract

MARAFIOTI, PABLO EZEQUIEL. Modeling of Plural Concordance Errors in Spanish as Foreign Language Using Dynamic Systems. Cuad. linguist. hisp. [online]. 2022, n.40, pp.1-.  Epub May 14, 2023. ISSN 0121-053X.  https://doi.org/10.19053/0121053x.n40.2023.14700.

The purpose of this paper is to model plural nominal concordance errors for a Spanish corpus of four learners of Spanish as a foreign language: SONIA (level A1/ A2), NATI (level B1), JAKO (level B2), MIRKA (level C1). In the framework of dynamical systems, it is postulated that "correct" and "incorrect" concordances constitute stable discrete attractors to which the continuous flow of the system converges in three simple simulation models: Lokta-Volterra, coordination game (evolutionary games theory), gradient ascent. As a result, the models better approximate the global error pattern than the session pattern. In the latter, the model based on Lokta-Volterra offers better performance for SONIA, NATI and JAKO. MIRKA obtains the worst performance in all the models, but its error decreases with the gradient ascent model. It is concluded that learning patterns are self-organized and emerge from the dynamics of concordance microsystems.

Keywords : simulation model; second language acquisition; dynamic systems; plural number.

        · abstract in Spanish | French | Portuguese     · text in Spanish     · Spanish ( pdf )