ABSTRACT
If I have to characterize the model that I will discuss, it will be in terms of two features that Kolers indicated he thought were contraindicated. First, this will be a bifunctional model, in the sense that it holds parafoveal vision to be extremely important in the skilled reading process. Second, skilled reading will be viewed as being "merely implicit" speech (in his terms), or more specifically, as being mediated by speech programs. As we shall see, implicit speech is a far more complex psychological function than a mouthing of phonemes and I think implicit speech itself is not "merely implicit speech."
