ABSTRACT
KLATT: Let me begin by spending one or two minutes with a few thoughts of my own on this subject. Some of you may have noticed a brief review paper in our Working Papers, VI, (Klatt, 1988), which discusses some of the recent models of speech perception. It is relatively easy to list candidate models, and I am not going to spend any time trying to define them or to apologize for those left off the list. The only observation I want to make at this point is that the most recent revised Motor Theory has changed from its original conception (a sort of analysis by synthesis module involving the articulators) to something that is not particularly well specified. One of my hopes this afternoon is that someone will say exactly what they think the revised Motor Theory would look like, if it were incorporated in a speech recognition device. Maybe most of you in the audience have a general idea of what that would look like, but perhaps not.
