Lexical Access: Another Theoretical Soupstone?
Mark S. Seidenberg · 2012
The lexical access way of thinking has set the agenda for research on lexical processing over the past two decades. Several theories have been proposed within this general framework (e.g., Coltheart , 1978; Forster, 1976; Morton, 1969). Although the theories differ in detail, they share the basic assumption tha t the units of representation in the mental lexicon correspond to the perceptual units in question, namely words. In Morton’s (1969) model, for example, each word is represented by an entry termed a logogen. Forster’s (1976) model consists of a master file, containing entries for all the words in a person’s vocabulary, and slave files, containing entries for their spellings, pronunciations, and meanings. Hence, knowledge of a word is represented in terms of entries tha t correspond to its different codes. Similarly, there is a class of models (see, e.g., Allport & Funnell, 1981; Pat terson & Shewell, 1987) in which knowledge of words is represented in terms of a set of orthographic, phonological, and semantic lexicons. The number of separate lexicons has proliferated, with recent models distinguishing between input lexicons used in perception and output lexicons used in production. For example, there are separate representations of the orthographic codes for words used in reading and spelling (Miceli, Silveri & Caramazza, 1985; Monsell, 1987). Finally, the models proposed by Glushko (1979) and McClelland and Rumelhart (1981) also assumed tha t each word has a corresponding entry in memory.