ABSTRACT

The conception of a series of cascading processing stages allows us to make specific predictions about skill interactions among components. Skilled reading may, in effect, represent the culmination point in the development of a powerful multiprocessor that can simultaneously analyze word structure, make lexical identifications, and process discourse structures and do all this in an integrated fashion. The priming effect of context was then the reaction time for reading words in context subtracted from that for similar words presented in isolation. Having established that there are good-poor reader differences in encoding of multiletter perceptual units, the question at issue is: What are the effects of this perceptual skill on a reader's subsequent decoding of orthographically regular words or pseudowords? Reader interprets words as lexical units that are referentially related to earlier text elements. The most salient variable indicative of lexical accessibility is, of course, word frequency.