Osamu Fujimura tried to bridge the gap between these two great papers. I am going to try to deepen the gulf. Whatever their authors' intentions, the gap between their theoretical positions can never be bridged. The reasons for this will become clear after we have examined a little more precisely what each paper says.

The Browman and Goldstein model has been described somewhat briefly before (Browman and Goldstein 1985, 1986); I am delighted to see that it is now becoming more articulated. Browman and Goldstein are giving us a valuable new theoretical framework, and presenting it in a way that is explicit and testable. They suggest that the phonological representation of a word should be partly in terms of a gestural score for articulatory features such as the manner and degree of constriction due to the tongue tip and that due to the tongue body. I want to consider what these tract variables (features?) can do from a linguistic point of view. First of all, are they sufficient? Consider, for example, the status of the jaw. In the current model, the jaw is not an articulatory feature that is part of what they call “the canonical phonological form.” So that we can understand the linguistic implications of the Browman and Goldstein view, we will consider whether it would be better if the movements of the jaw were in fact also given by a separate gestural tier in their phonological representation.

There are many individual differences in the way that people coordinate jaw and tongue body movements. Some speakers control the position of the tongue essentially by jaw movements when making the vowels in “heed, hid, hayed, head, had.”