ABSTRACT
In Chapter 8 the competence-performance distinction was incorporated into the naive model of Chapter 6. The purpose of this theoretical extension was to help account for some serious false predictions on the part of the original model. The following question naturally arises: When should a model be salvaged in this way? How can we tell if the wrong model is being protected? When are we rationalizing the false predictions of a basically sound theory, and when are we simply on the wrong track? There seems to be no easy answer to these questions, although “simplicity” or “elegance” are sometimes invoked.
