Using Visual Information to Train Foreign‐Language Vowel Production*

The present feasibility study explored the use of visual information for training vowel production in a foreign language (L2). An optoelectronic glossometer was used to measure tongue‐palate distances at four locations along the hard palate and to provide visual feedback specifying tongue targets for English /i,I,æ/ and /α/. Evaluations performed during 10 days prior to training showed that the native Spanish subject examined consistently neutralized the tongue position difference between English /i/‐/I/ and /æ/‐/α/. The same thing occurred when she attempted to imitate vowels spoken by an English native speaker. Physiological, acoustic, and perceptual tests showed that the subject produced a difference between /i/‐/I/ after only ten minutes of visual articulatory modeling and shaping. A similar improvement was not noted for /æ/‐/α/, perhaps because the difference in pharyngeal width that distinguishes these vowels could not be shown directly using the glossometer. Additional work will be needed to determine whether the /i/‐/I/ change was due to the visual feedback and whether greater—and sustained—changes in vowel production can be effected using visual feedback if more training is administered.