20 - Toward a model of articulatory control: comments on Browman and Goldstein's paper

Summary

Browman and Goldstein provide an outline of a new quantitative model of the temporal organization of speech that deviates basically from the classical model of concatenation and coarticulation (see Fujimura 1987 for some relevant discussion of this issue). Admittedly, some of the current specifications of their model are vague, and further work is much needed. Also, I am not quite convinced that the model they propose will work, after all. My own approach is rather different (Fujimura 1987). Nevertheless, I welcome this proposal with enthusiasm and excitement. I share their sense of importance of the goal of establishing a quantitative model of temporal organization, and whoever's idea may turn out to be correct, it will be great to see some solution to this problem of our strong concern. The problem is very difficult to solve, but it has to be solved to be able to understand the essence of speech organization, to arrive at an adequate framework of phonological description, and also to apply such basic linguistic knowledge for true breakthroughs in speech technology.

My first comment is that it is not clear to me what task dynamics actually contributes to the description proposed here for relating articulatory gestures to each other. I can see that the current specifications of the model seem consistent with the idea of task dynamics, but I still have to be convinced that further specifications which will need to be given to make the model adequate for describing the speech organization we observe, do not cause any theory-internal inconsistency.