Unraveling the Role of Statistical Learning in Second Language Learning and Processing
Final Report Abstract
Extensive research has shown that statistical learning (SL) - succinctly defined as the discovery of structure by way of statistical properties of the input - is a fundamental mechanism that shapes basic aspects of perception and action. In the field of language learning, SL has been shown to boost the acquisition of knowledge across all components of the linguistic system, from the acquisition of properties of speech as well as the graphotactic and morphological regularities of written words to complex grammatical structures. Moreover, there is growing evidence of a tight coupling between individual differences in SL abilities and variability in native language learning and processing, in both child and adult populations. However, while this research has been expanding rapidly in recent years, much less is known about the role of SL in second-language (L2) learner populations. The initial phase (P1) of the current project has significantly advanced our understanding of SL and its connection to Second Language Learning and Processing (L2LP). This was achieved through a systematic exploration of the extent to which L2 learners can assimilate linguistic knowledge by analyzing statistical regularities inherent in the L2 input. The outcomes of P1 have laid the groundwork for a more comprehensive and integrated approach, setting the stage for addressing pivotal theoretical and methodological questions in the subsequent phase (P2). The results from P2 have illuminated intricate interconnections between SL abilities and diverse factors related to L2 experience, as well as cognitive-affective individual differences, delineating their relative impact on L2LP. On a methodological level, innovative processing-based measures were introduced to assess SL abilities, and these were effectively aligned with L2 performance measures based on ecologically valid data. The project synthesized empirical evidence from a series of corpus-based and behavioral experimental studies, offering a profound understanding of the intricate dynamics between SL and inter-individual differences factors influencing L2 production and comprehension. By integrating SL and implicit learning paradigms, the project investigated both the phenomenological aspects and learning consequences of SL knowledge. Furthermore, it underscored the importance of harnessing big data, natural language processing, and machine/deep learning, demonstrating their combined efficacy in gaining new insights and fostering a more profound understanding of language learning on a large scale. Collectively, the findings from this project have provided valuable insights into the fundamental mechanisms underlying language learning and processing. The findings on individual differences are of central importance for contemporary theories on the acquisition, processing and use of language and have direct relevance for L2 instruction.
Publications
-
IMPLICIT LEARNING IN THE CROWD. Studies in Second Language Acquisition, 39(4), 711-734.
Kerz, Elma; Wiechmann, Daniel & Riedel, Florian B.
-
Effects of Statistical Learning Ability on the Second Language Processing of Multiword Sequences. Lecture Notes in Computer Science, 200-214. Springer International Publishing.
Kerz, Elma & Wiechmann, Daniel
-
L2 Processing Advantages of Multiword Sequences: Evidence from Eye-Tracking. Proceedings of the Joint Workshop on Multiword Expressions and WordNet (MWE-WN 2019), 60-69. Association for Computational Linguistics.
Kerz, Elma; Heilmann, Arndt & Neumann, Stella
-
Tuning to Multiple Statistics: Second Language Processing of Multiword Sequences Across Registers. In Proceedings of the 41st Annual Meeting of the Cognitive Science Society, Montreal, Canada (pp. 546-552)
Kerz, E., Wiechmann, D. & Christiansen, M. H.
-
Becoming Linguistically Mature: Modeling English and German Children’s Writing Development Across School Grades. Proceedings of the Fifteenth Workshop on Innovative Use of NLP for Building Educational Applications, 65-74. Association for Computational Linguistics.
Kerz, Elma; Qiao, Yu; Wiechmann, Daniel & Ströbel, Marcus
-
The Relationship Between First and Second Language Writing: Investigating the Effects of First Language Complexity on Second Language Complexity in Advanced Stages of Learning. Language Learning, 70(3), 732-767.
Ströbel, Marcus; Kerz, Elma & Wiechmann, Daniel
-
The role of verbal working memory in L2 sentence comprehension. Journal of Second Language Studies, 3(1), 1-30.
Kerz, Elma & Wiechmann, Daniel
-
Understanding the dynamics of second language writing through keystroke logging and complexity contours. In Proceedings of the Twelfth Language Resources and Evaluation Conference (pp. 182-188)
Kerz, E., Pruneri, F., Wiechmann, D., Qiao, Y. & Ströbel, M.
-
Automated classification of written proficiency levels on the CEFR-scale through complexity contours and RNNs. In Proceedings of the 16th Workshop on Innovative Use of NLP for Building Educational Applications (pp. 199-209)
Kerz, E., Wiechmann, D., Qiao, Y., Tseng, E. & Ströbel, M.
-
Assessing linguistic complexity and register flexibility in advanced second language learners. Register Studies, 4(1), 55-90.
Kerz, Elma; Neumann, Stella & Niemietz, Paula
