The Institute for Japanese Studies presents:
How effective is training on Japanese sound learning? An AI-assisted meta-analysis of training studies
Seth Goss
Emory University
Abstract: Vowel and consonant length contrasts (e.g., kado ‘corner’ vs. kaado ‘card’) along with pitch accent are two areas of Japanese phonology that many second language (L2) learners find challenging to acquire. Both length contrasts and pitch are important not only for learners’ own listening comprehension, but also the intelligibility of their spoken Japanese. Yet instructors understandably have little time to target these features, given the constraints on classroom hours. This has prompted researchers to develop various training tasks to help learners both perceive and produce length and pitch. However, no studies have taken a step back to ask: “How effective is training on the learning of these speech sounds?” In this talk I report on a meta-analysis of primary studies on the effectiveness of training on L2 Japanese sound learning. I will also show how I used a “human in the loop” AI tool to assist with the literature screening.
Seth Goss is Associate Professor of Japanese and Director of the East Asian Studies Program in the Department of Russian and East Asian Languages and Cultures (REALC) at Emory University. He is also a core faculty member in Emory’s Linguistics Program. His research focuses on second/third language sound learning, individual difference measures, and vocabulary learning. He teaches courses on Japanese linguistics and bilingualism studies, in addition to language classes. He is the 2020 recipient of the Michael Sharwood Smith Award for best article published in the journal Second Language Research. He has also published his research in journals including Applied Psycholinguistics, Studies in Second Language Acquisition, and Journal of Japanese Linguistics. He received his PhD in Japanese Linguistics from DEALL in 2015.