Phonological encoding is not contingent on semantic feature retrieval: An electrophysiological study on object naming

In the present study, the authors examined with event-related brain potentials whether phonological encoding in picture naming is mediated by basic semantic feature retrieval or proceeds independently. In a manual 2-choice go/no-go task the choice response depended on a semantic classification (animal vs. object) and the execution decision was contingent on a classification of name phonology (vowel vs. consonant). The introduction of a semantic task mixing procedure allowed for selectively manipulating the speed of semantic feature retrieval. Serial and parallel models were tested on the basis of their differential predictions for the effect of this manipulation on the lateralized readiness potential and N200 component. The findings indicate that phonological code retrieval is not strictly contingent on prior basic semantic feature processing. The ability to recognize and name objects from vision is a seemingly effortless, highly practiced everyday skill. For example, after seeing a big gray creature with large ears and a trunk we can easily produce the name elephant at high speed. Visual perception also triggers access to rich sources of semantic information stored in memory. Attributes like is an animal, is a herbivore, lives i