| Number | 1135 |
|---|---|
| Year | 1999 |
| Drawer | 21 |
| Entry Date | 12/03/1999 |
| Authors | Honorof, D.N. |
| Contact | |
| Publication | Proceedings of the XIVth International Congress of Phonetic Sciences, (pp.1141-1144). San Francisco, CA. |
| url | http://www.haskins.yale.edu/Reprints/HL1135.pdf |
| Abstract | Magnetometrically transduced articulator position data were collected for pre-vocalic American English /r/ and /l/ as spoken by four adult Japanese learners of English under native-speaker imitation and non-imitation conditions. Japanese has only one liquid phoneme, therefore imitation of these contrasting and reportedly difficult English sounds was expected to decrease token-to-token variability and to increase overall distinctiveness of sets of articulator positions for the two members of the contrast. Stepwise discriminant analyses and variability testing proved support for two claims. 1) The visual modality may be relied upon during imitation when the information it provides is especially salient. 2) Imitation of a native-speaker model facilitates articulatory control over challenging L2 speech contrasts in inverse relation to previous mastery of control for the individual targets. |
| Notes |