Resumen de Learning grammatical constructions from audio-visual input
This doctoral dissertation explores the effects of prolonged exposure to audio-visual input on learning a variety of English L2 grammatical constructions. The thesis is organised around three interconnected studies. The first study compares the effects of captioned and uncaptioned audio-visual input on learning grammar, and investigates whether this learning depends on such learner-related factors as L2 proficiency level, working memory capacity, and foreign language learning aptitude. The second study implements an additional captioning condition and compares grammar gains between the three captioning modes: Unenhanced captions, textually enhanced captions, and no captions. It also assesses the effects of three grammatical construction learnability factors: Construction type, frequency, and recency of occurrence. The third study turns to the viewers’ perspective of learning from audio-visual input and addresses the participants’ feeling of learning from the intervention, and whether their extramural L2 television viewing preferences and viewing strategies changed over the period of the intervention.
A total of 141 participants with various proficiency levels (from A1 to C2) watched ten full-length TV series episodes with captions, with textually enhanced captions (Study 2), or without captions over a period of five weeks. The study targeted 27 frequently occurring grammatical constructions categorized as fully-schematic, partially-filled, or fully-filled. The design included a pre-test, an immediate post-test, and a delayed post-test, along with pre-course and post-viewing questionnaires on participants’ L2 viewing habits.
Participants’ individual differences such as English proficiency level, working memory capacity, and foreign language learning aptitude were also measured.
The results of the first study indicated that all participants significantly improved their knowledge of the target grammatical constructions, with the captions group having significantly higher gains than the no captions group. In terms of the individual differences, English proficiency had a mediating role with the intermediate level group outperforming the elementary group. The working memory capacity and language learning aptitude scores (LLAMA F) had a significant effect on the learning gains of the no captions group. Higher levels of these cognitive individual factors facilitated learning without captions, suggesting that captions can level the playing field while viewing L2 audio-visual materials.
The results of the second study indicated mixed effects of captioning. Textually enhanced captions – a more salient condition – led to immediate learning outcomes, while unenhanced captions resulted in higher long-term effects. This finding suggests that in order to obtain effective learning conditions from textually enhanced captions, a limit to the amount of different textually enhanced constructions presented in the input should be set. In general, unenhanced captions - already available on the most streaming platforms – appear sufficient for successful grammatical constructions learning. As regards the type of constructions and frequency, fully-filled constructions were learnt the least by all the groups. It seems that constructions without any variation in the input were more difficult to uptake. Frequency of the target constructions occurrence did not have a significant effect on grammatical constructions learning of intermediate and advanced level participants. It is possible that lower language proficiency viewers are more sensitive to the frequencies in the audio-visual input.
The third study, exploring the learners’ perspectives, found that the participants perceived vocabulary and expressions as the most learnt language features. Concerning the preferred viewing mode outside of the classroom (with L1 subtitles, with L2 captions, without captions or subtitles), a significant shift was observed. All participants, regardless of their proficiency and intervention viewing group, opted for watching less with L2 captions. The elementary level participants found that viewing without any native language support was too challenging for leisure viewing, while the intermediate and advanced level students became familiar with the audio-visual input and turned into confident viewers. Finally, the participants also demonstrated a significant drop in applying viewing strategies, which could be attributed to them turning off L2 captions and becoming viewers rather than learners.
This doctoral dissertation extends the known benefits of sustained exposure to audio-visual input for L2 vocabulary learning and comprehension, and goes beyond by providing evidence for the potential of captioned L2 television for grammar learning.
