ABSTRACT
A cross-modal lexical priming procedure was used in three experiments that investigated across three time delays (0,400, and 1000ms) the question of whether there are ‘core’ meaning components in lexical interpretation. The results suggested that beyond the initial spread activation of associates, core features of a concept tend to be instantiated in the representation of a sentence that mentions the corresponding word, regardless of their contextual relevance. However, weak features tend to be instantiated only in relevant contexts. It is concluded that initial lexical access is context independent and that core features function as default meaning components of a word.
