Simner and Smyth (1998) propose that anaphoric lexical access (ALA) occurs at an anaphor and targets the lexical entry (specifically, the lemma) of the antecedent. Since the word frequency effect (e.g., Rubenstein et al., 1970) resides at the lexeme (Jescheniak & Levelt, 1994) Simner and Smyth predict, and subsequently illustrate, that ALA exhibits no frequency effect. A problem arises, however: if ALA does not access the lexeme, how do we account for phonological priming at anaphor sites (e.g., Tanenhaus et al., 1985)? We claim that this is the result of "incidental" lemma-to-lexeme activation. Furthermore, we argue that since lexeme activation is not crucial to anaphor comprehension, anaphor reading times indicate lemma search times only (therefore there is no frequency effect). An experiment is presented demonstrating that lemma-access during ALA can cause incidental lexeme activation without invoking a frequency effect.
Copyright 1999 Academic Press.