Enhanced input and enriched context to improve the acquisition of the Spanish grammatical gender assignment and agreement: A preliminary study

Abstract

Research in Second Language Acquisition has diagnosed extensively how the acquisition of a language's grammatical gender system poses problems to learners of a second language while it is acquired effortlessly by native speakers. In addition, the main issue that has been addressed by the literature concerns assignment alone, thus not taking into consideration the full scope of the grammatical gender system. Research on types of learning has shown that it is possible to learn incidentally, namely without noticing, which allows for more efficiency time-wise. However, several factors modulate both the acquisition of a language's grammatical gender learning and types of learning, the main factor being proficiency. With the goal to facilitate the acquisition of the Spanish grammatical gender for learners, this study puts forward a pedagogical approach that consists of enhanced and enriched input. Four treatment groups were created in a crossed design with two groups with enhanced input (bold font) and two without it. One group with enhanced and one unenhanced had enriched context (article-noun-adjective combination) and one did not (article-noun combination). Participants were exposed to a PowerPoint presentation that presented 30 novel nouns that were controlled by grammatical gender (masculine or feminine) and noun ending (canonical, non-canonical, and exceptions) combined with 40 novel adjectives (in the article-noun-adjective condition, of which 20 were variable and 20 were invariable). Results showed that treatments did not render significant differences and that proficiency was the best predictor for accuracy in gender assignment and agreement. Masculine nouns showed significantly higher accuracy rates with respect to feminine nouns and nouns with a canonical ending showed higher accuracy rates with respect to non-canonical and exceptional endings. Results in the delayed post-test show that there is a significant improvement in performance for canonical noun endings, while exceptional endings decrease significantly and non-canonical endings remain the same. No differences are observed between tests regarding gender of the nouns. These results provide tentative evidence for Ullman's (2001) Declarative/Procedural model since results would suggest that the unmarked noun endings were proceduralized, but not other features. It is also argued that these learners show improvement depending on their proficiency, thus being capable of ultimate attainment. The results suggest that the errors made by learners are the result of features being reassembled in the learners' internal linguistic system.