Use this URL to cite or link to this record in EThOS:
Title: Perception and production of English /r/-/l/ by adult Japanese speakers
Author: Hattori, K.
ISNI:       0000 0004 2726 7485
Awarding Body: University College London (University of London)
Current Institution: University College London (University of London)
Date of Award: 2010
Availability of Full Text:
Access from EThOS:
Access from Institution:
This dissertation examines perception and production of English /r/-/l/ by adult Japanese speakers. This programme of research is organized into three sections, termed Study 1, Study2, and Study 3. The first study examined whether category assimilation between English /r/-/l/ and Japanese /ɾ/ was predictive of /r/-/l/ identification accuracy using an individual difference approach. Japanese speakers were assessed in terms of /r/-/l/ identification and assimilation of English /r/-/l/ into Japanese /ɾ/, /r/-/l/ production, and perceptual best exemplars for /r/, /l/, and /ɾ/. The results demonstrated that, although Japanese speakers strongly assimilated /l/ to /ɾ/, category assimilation was not predictive of English /r/-/l/ identification accuracy, and that only Japanese speaker’s representations for F3 in /r/ and /l/ was predictive of /r/-/l/ identification ability. The second study similarly took an individual difference approach and examined whether there is a relationship between perception and production of /r/-/l/ measuring perception accuracies (i.e., identification, discrimination, and perceptual best exemplars) and production accuracies (i.e., acoustic measurements, and recognition accuracy by English speakers). The results demonstrated that perception and production of /r/-/l/ were moderately related. However, not all aspects of /r/-/l/ perception were incorporated into /r/-/l/ production. The third study examined whether one-to-one pronunciation training leads to improvement in production and perception of English /r/-/l/ using a multipronged approach (i.e., explicit instructions, real-time spectrograms, and feedback with signal-processed versions of their own productions). The results demonstrated that Japanese speakers could be trained to produce native-like English /r/-/l/, improving to the point that they approached a 100% accuracy ceiling in terms of how well native speakers could identify their productions. However, the training did not improve their English /r/-/l/ perception at all.
Supervisor: Not available Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID:  DOI: Not available