Search Papers | Poster Sessions | All Posters
Poster C64 in Poster Session C - Friday, August 9, 2024, 11:15 am – 1:15 pm, Johnson Ice Rink
L1 Lexical Semantic Knowledge Shapes Statistical Word Learning
Benjamin Zinszer1 (), Vic Tianlan Wen1, Adrian Ferguson1, Shinz Jo Ooi1, Shuchang "Coco" Liu2; 1Swarthmore College, 2Haverford College
Learners’ performance on cross-situational statistical learning (CSSL) paradigms generally decreases when novel word-object input diverges from simple 1-to-1 mappings. Natural language learning reflects a similar limitation: Bilinguals often rely on direct translation of words from the first language (L1) to the second (L2) before learning L2-specific lexical semantic structures. We looked at the interface between learners’ existing L1 lexical semantics and input from a brief CSSL task structured to simulate cross-language differences. When mappings violated English object naming norms, English monolinguals learned them more slowly than English-consistent mappings. Accuracy on a 9-AFC test phase correlated with the English naming norms for the depicted objects.
Keywords: statistical learning lexical semantics language learning cross-linguistic