Accurate mirroring in (haptic-integrated) pronunciation teaching

Clip art: Clker

Clip art: Clker

Mirroring, having learners move along with a model, is a common technique in pronunciation teaching, especially at more advanced levels such as this by Goodwin at UCLA. There are any number of applications of the concept, for various purposes. In EHIEP work, mirroring figures in prominently from the beginning. As noted in previous posts, some highly visual learners find imprecise modelling by the speaker being mirrored to be very disconcerting. For example, one pedagogical movement pattern (PMP) involves moving the left hand across the visual field in an ascending motion as a rising intonation contour is spoken. For sometime we have been looking at the possibility of using avatars that would perform perfectly precise PMPs in new versions of the haptic videos to compensate for the fact that a human model (namely me on the current videos) cannot possibly be consistent enough to satisfy the few most radically visual learners. Research by Thomaz at Georgia State University seems to suggest that the only way to do that with robotic models--would be to build in "human-like" variability of motion into the repetitions of PMPs in training. In other words, the slight differences in the track of the gestural patterns is essential to creating a sufficiently engaging model to effectively keep subjects' attention. Rats. Better go back to figuring out both how to be more "humanly" precise in modelling PMPs and developing techniques that will assist the "visually-challenged" in loosening up a bit. Figuring out exactly what acceptable deviations from ideal PMPs are is, in principle, doable, of course. Just a matter of studio time and field testing. Keep in touch. 

Phonetic (or phonemic) gesture revisited (in the classroom)

Clipart: Clker

Clipart: Clker

In the development of our understanding of speech perception, one of the terms used by some researchers was "phonetic gesture." It essentially referred to the process by which sounds are perceived--the articulatory, not the acoustic properties. The key question was just how much one's ability to articulate a sound determined ones's ability to perceive it. What subsequent research has shown is that it is a mixed bag; the relationship between external properties of sound and our internal processing of it is very complex and developmental as well. In short, ongoing perception of speech turns out to be more a matter of our conceptual systems "expectations" than it is with the actual physical properties of what we hear. That is not to say that the felt sense of the bodily "mechanics" is not important and cannot contribute both to understanding and learning. I like the term, phonetic gesture, as relating to the somatic, physical side of sound production and perception. In our work, a better application of that idea might be "phonemic gesture," that is pedagogical movement patterns that represent key meaningful units of sound within English, including vowels, rhythm patterns, stress assignment and intonation contours. As noted earlier, one of my first, informal research studies was to sit in classes of my colleagues and take notes on the use of gesture they used to accompany pronunciation instruction. Those observations got me started on this line of thinking about 20 years ago. That language instructors adapt gesture for many purposes was the subject of this research by Stam and Teller. (Their work is reported in other publications as well.) One interesting finding was the expansion of the range and depth of field of motion of gesture used " . . . an equivalent of shouting in gesture form." So, what is your current pedagogical "phonemic gesture inventory?" What do you mean?