Bibliografía - Revista Española de Lingüística Aplicada

Extended oral production has seldom been used to explore adjectival and verbal agreement in L2 Spanish. This study examines oral narrations to compare the agreement behavior, speech rates, and patterns of errors of highly proficient Spanish heritage and L2 learners (early and late bilinguals, respectively), whose L1 is English, with those of native controls. Although both bilingual groups displayed high agreement accuracy scores, only the early bilinguals performed at or close to ceiling. In addition, the L2 learners spoke significantly more slowly than the heritage and native speakers, who displayed similar speech rates. Explanations accounting for the differences in speech rates and agreement accuracy include age of acquisition of Spanish, syntactic distance between a noun and its adjective, and task effects. All of these factors favored the early bilinguals, enhancing their advantages over L2 learners. Findings suggest that the integrated knowledge and automatic access needed for native-like attainment in agreement behavior in extended oral production is more easily achievable by early than by late bilinguals.

Texto completo pdf icon

This article presents the Spanish version of the Computer-Assisted Listening and Speaking Tutor (CALST), an online platform that can be used to complement pronunciation training in the classroom. The Spanish CALST offers listening, speaking, and spelling exercises for Northern-Central Peninsular Spanish as an L2. Exercises are tailored by an automatic comparison with the learner’s native language based on a database of phonetically specified phoneme inventories for over 500 languages, with the result that learners with different L1s are exposed to different exercises adapted to their specific needs. In this article, we present a description of the exercises as well as the criteria used to develop Spanish content for CALST. We discuss the limitations of the platform, the logging of user results as a partial solution to these limitations, and the possible future use of the logged data to increase our understanding of L2 acquisition.