Eye-tracking shows that digital training improves L2 learning of stress-suffix associations
Author ORCID Identifier
https://orcid.org/0009-0004-6298-2596
Document Type
Paper Presentation
Presenter Language
English
Research Area
Psycholinguistics
Location
MBSC Gallery Room 308
Start Date
19-10-2024 12:00 PM
End Date
19-10-2024 12:30 PM
Abstract
Prosody assists us in parsing continuous speech and recognize words. For instance, native speakers use prosodic information such as tone and stress to predict word endings (see Roll et al., 2022, for a review). In contrast, beginning L2 learners struggle perceiving L2 prosodic cues used differently in their L1 (e.g., Sagarra & Sagarra, 2018). Prosody is often neglected in L2 classrooms (Gosselke Berthelsen et al., 2023), but Schremm et al. (2017) found that digital gaming improved L1 non-tonal beginning learners of a tonal L2 acquire L2 tone-suffix associations. Since this improvement could be due to practice effects, tone learning, or only segmentals, we asked L2 learners to complete an eye-tracking pretest/posttest that assessed their ability to apply what they learned in the digital game to new instances, we examined another segmental—lexical stress (oxytone, paroxytone)—, and we explored a suprasegmental—syllabic structure (CV, CVC).
Twenty L1 English-L2 Spanish beginning learners played a digital game for 15 minutes per day for ten days. Participants heard the beginning of a sentence cut at the verb’s initial syllable (la amiga pes- ‘the friend weigh-”), selected a present or a past suffix, and received feedback. Before and after training, participants completed an eye-tracking task assessing their ability to use the stress of a Spanish disyllabic verb’s first syllable to predict the verb’s suffix. None of the target verbs in the eye-tracking tasks were part of the training.
Generalized Additive Mixed Models (GAMMs) revealed that playing the digital game increased accuracy and decreased reaction times. Furthermore, before playing the game, learners did not use prosodic information to make stress-suffix associations, but after playing the game, learners anticipated word endings in the CVC oxytone condition, which represents the less frequent pattern. This result aligns with previous studies showing that predictive processing is facilitated when cues are associated with fewer lexical competitors. These findings suggest that training can help L2 learners gain more nativelike predictive processing. The data support models positing that L2 learners can acquire segmentals and suprasegmentals different from their L1 and that prosody influences lexical access Keywords: second language learners; prosody; training, prediction.
References
Gosselke Berthelsen, S., & Roll, M. (2023). Computer-aided L2 prosody acquisition and its potential in second language learning. ASLA:s skriftserie, 30, 157-182. https://lup.lub.lu.se/record/30c68231-a9fc-4e16-bc20-cdc015ebd7e8
Roll, M. (2022). The predictive function of Swedish word accents. Frontiers in Psychology, 13, 910787. https://doi.org/10.3389/fpsyg.2022.910787
Sagarra, N., & Casillas, J. V. (2018). Suprasegmental information cues morphological anticipation during L1/L2 lexical access. Journal of Second Language Studies, 1(1), 31-59. https://doi.org/10.1075/jsls.17026.sag
Schremm, A., Hed, A., Horne, M., & Roll, M. (2017). Training predictive L2 processing with a digital game: Prototype promotes acquisition of anticipatory use of tone-suffix associations. Computers & Education, 114, 206-221. https://doi.org/10.1016/j.compedu.2017.07.006
Abstract with figures
Eye-tracking shows that digital training improves L2 learning of stress-suffix associations
MBSC Gallery Room 308
Prosody assists us in parsing continuous speech and recognize words. For instance, native speakers use prosodic information such as tone and stress to predict word endings (see Roll et al., 2022, for a review). In contrast, beginning L2 learners struggle perceiving L2 prosodic cues used differently in their L1 (e.g., Sagarra & Sagarra, 2018). Prosody is often neglected in L2 classrooms (Gosselke Berthelsen et al., 2023), but Schremm et al. (2017) found that digital gaming improved L1 non-tonal beginning learners of a tonal L2 acquire L2 tone-suffix associations. Since this improvement could be due to practice effects, tone learning, or only segmentals, we asked L2 learners to complete an eye-tracking pretest/posttest that assessed their ability to apply what they learned in the digital game to new instances, we examined another segmental—lexical stress (oxytone, paroxytone)—, and we explored a suprasegmental—syllabic structure (CV, CVC).
Twenty L1 English-L2 Spanish beginning learners played a digital game for 15 minutes per day for ten days. Participants heard the beginning of a sentence cut at the verb’s initial syllable (la amiga pes- ‘the friend weigh-”), selected a present or a past suffix, and received feedback. Before and after training, participants completed an eye-tracking task assessing their ability to use the stress of a Spanish disyllabic verb’s first syllable to predict the verb’s suffix. None of the target verbs in the eye-tracking tasks were part of the training.
Generalized Additive Mixed Models (GAMMs) revealed that playing the digital game increased accuracy and decreased reaction times. Furthermore, before playing the game, learners did not use prosodic information to make stress-suffix associations, but after playing the game, learners anticipated word endings in the CVC oxytone condition, which represents the less frequent pattern. This result aligns with previous studies showing that predictive processing is facilitated when cues are associated with fewer lexical competitors. These findings suggest that training can help L2 learners gain more nativelike predictive processing. The data support models positing that L2 learners can acquire segmentals and suprasegmentals different from their L1 and that prosody influences lexical access Keywords: second language learners; prosody; training, prediction.
References
Gosselke Berthelsen, S., & Roll, M. (2023). Computer-aided L2 prosody acquisition and its potential in second language learning. ASLA:s skriftserie, 30, 157-182. https://lup.lub.lu.se/record/30c68231-a9fc-4e16-bc20-cdc015ebd7e8
Roll, M. (2022). The predictive function of Swedish word accents. Frontiers in Psychology, 13, 910787. https://doi.org/10.3389/fpsyg.2022.910787
Sagarra, N., & Casillas, J. V. (2018). Suprasegmental information cues morphological anticipation during L1/L2 lexical access. Journal of Second Language Studies, 1(1), 31-59. https://doi.org/10.1075/jsls.17026.sag
Schremm, A., Hed, A., Horne, M., & Roll, M. (2017). Training predictive L2 processing with a digital game: Prototype promotes acquisition of anticipatory use of tone-suffix associations. Computers & Education, 114, 206-221. https://doi.org/10.1016/j.compedu.2017.07.006