Word Perception II: Word Identifi cation in Text
In Chapter 3 we introduced methods for studying isolated words and some basic issues in how printed words are encoded. From the research we discussed, several conclusions emerged: (1) encoding of shorter words occurs through a process in which the letters are processed in parallel; (2) conversion to sound is defi nitely involved in accessing the meaning of a word for skilled readers, although there is some controversy about how important or universal that involvement is; (3) for longer words (especially those with more than one morpheme) this parallel processing is likely to break down and such words are likely to be processed in more than one chunk; (4) word encoding is relatively automatic. (We will have more to say about the last issue later in this chapter when we discuss whether more than one word is processed at a time.) We also asserted in Chapter 3 that the conclusions drawn from those studies were largely true when words were in sentence context. Thus it would seem like a good place to start this chapter by documenting that most, if not all, of these conclusions have been borne out with eye movement data before plunging into the newer questions that have become the focus of much of the word-processing literature in the last 20 years.