ABSTRACT
In order to understand how a word is read for meaning, we need to know how a reader proceeds from the printed representation of a word to the word’s entry in the reader’s internal lexicon, where the word’s meaning is stored. This raises two principal questions: what is the code in which the word is represented when this process, lexical access, is being carried out, and what is the procedure by which this representation is used to find the word’s entry in the lexicon. The lexical-decision task is a suitable one for the investigation of these questions. Two experiments using this task are reported. In one, it was found that a letter string’s similarity to English words influenced the “no” response latency, but not the “yes” response latency, and it is argued that this result favors the view that lexical access is “direct,” rather than requiring search. The other experiment showed that a nonword’s phonological properties influenced the time taken to say “no” to it. Thus phonological encoding is occurring in these experiments. It remains to be shown, however, that this is any more than an epiphenomenon; neither these findings, nor those of previous investigators, compel us to abandon the view that skilled reading of single words proceeds solely by making use of visual representations of printed words.
