Yoh'ichi Tohkura's research while affiliated with National Institute of Informatics and other places

Publications (3)

Article
Full-text available
The present study used magnetoencephalography (MEG) to examine perceptual learning of American English /r/ and /l/ categories by Japanese adults who had limited English exposure. A training software program was developed based on the principles of infant phonetic learning, featuring systematic acoustic exaggeration, multi-talker variability, visibl...
Article
Full-text available
Linguistic experience alters an individual's perception of speech. We here provide evidence of the effects of language experience at the neural level from two magnetoencephalography (MEG) studies that compare adult American and Japanese listeners' phonetic processing. The experimental stimuli were American English /ra/ and /la/ syllables, phonemic...

Citations

... Across domains of visual perception, auditory perception, motor learning, language, inductive reasoning, problem solving, and computational modeling, a general observation is that increased input variability may come at a cost of initially hindering learning but often show subsequent benefits in generalization (Raviv et al., 2022). Although a significant amount of work has been devoted to understanding the cognitive and neural mechanisms supporting speech learning (e.g., De Diego-Balaguer & Lopez- Barroso, 2010;Zhang et al., 2009), much less work has considered how individual learners' cognitive abilities may influence the efficacy of speech training in terms of perceptual generalization, transfer of learning to production and long-term retention, as perceptual learning does not solely depend on the nature of exposure, but also learner ability to cope with stimulus variability. ...
... ularly true when the task at hand is more challenging. The acoustic exaggerations on irrelevant dimensions for L2 may impede this process due to the "Native Language Neural Commitment" that prioritizes the allocation of perceptual attention and processing resources to optimize efficient speech categorization in service of L1 phonology instead of L2 (Zhang et al., 2005(Zhang et al., , 2009. Given that we did find a long-term effect of our multipletalker training on the synthetic phoneme identification task, it is possible that the identification of naturally-produced words was more difficult because it might have incurred an additional load for processing lexical information (Escudero et al., 2008) and semantic content (Guion & Pederson, 2007). ...